PERPLEXITY AI- JUNE TO JULY 2026

[6/13, 11:56 PM] OUR DAILY WISDOM: Meet Perplexity

Your team lead asks you to research three cities for a team offsite by the end of the week. You open Google. Forty minutes later, you have 12 tabs, three conflicting blog posts, and still no clear answer to bring to the meeting.

That's exactly the problem Perplexity was built to solve. By the end of this lesson, you'll know what it does, how it works, and when to use it.

[6/24, 11:49 PM] OUR DAILY WISDOM: Meet Perplexity

That's exactly the problem Perplexity was built to solve. By the end of this lesson, you'll know what it does, how it works, and when to use it.

[6/24, 11:49 PM] OUR DAILY WISDOM: Before we go further, what is one common challenge in online research?

Search engines struggle with complex questions

There isn't enough information available online

There's a lot of information, but it's not synthesized into a clear answer

Correct answer

Right. The real challenge today usually isn't access to information — it's making sense of it. You often open multiple tabs, compare sources, and connect the dots yourself.

[6/24, 11:51 PM] OUR DAILY WISDOM: There's a lot of information, but it's not synthesized into a clear answer

Correct answer

Right. The real challenge today usually isn't access to information — it's making sense of it. You often open multiple tabs, compare sources, and connect the dots yourself.

What Is Perplexity

Perplexity is an AI-powered search engine that, unlike Google, reads its sources and hands you one synthesized answer — with citations attached to every claim so you can verify what you're looking at before acting on it. Let's see it in action.

Your First Search

Ask Preplexity a question about the team's off-site.

Your First Search

Ask Preplexity a question about the team's off-site.

Perplexity

Best European cities for a team offsite under $2000 per person

Budapest, Prague, and Riga top the list for team off-sites under $2000/person (3 days/2 nights, 10-person group from Western Europe).

Quick Costs (Per Person):

— Budapest: $550-900 (cheap flights $100-200, hotel $150-250, spas/activities $200).

— Prague: $530-880 (walkable sites, beer tastings, meeting spaces).

— Riga: $430-740 (UNESCO Old Town, parks, escape rooms).

Why These? Affordable Eastern Europe hubs with co-working venues, team activities (cruises, Olympics games), and logistics support — perfect value vs. pricier Lisbon/Athens.

Sources: Wise, Team-Retreats

Amazing!

You're right on track with your approach

[6/24, 11:52 PM] OUR DAILY WISDOM: Compare this to opening Google for the same question. What did Perplexity do that Google doesn't?

It searched more websites than Google would

It personalized results based on your location

It read multiple sources and gave one synthesized answer

Correct answer

Right. The volume of sources isn't the difference — Perplexity did the reading and condensing for you.

[6/24, 11:53 PM] OUR DAILY WISDOM: So Perplexity reads, synthesizes, and shows its work all in one step. But if that sounds familiar, you might be thinking of ChatGPT.

[6/24, 11:54 PM] OUR DAILY WISDOM: Perplexity vs. ChatGPT

Perplexity might look similar to ChatGPT at first — you type a question and get an answer.

Both can search the web, but that's where the similarity ends. For Perplexity, live web search is the default — every answer comes with citations. For ChatGPT, web search is optional. It's built for conversation, writing, and creative tasks, not for sourced, real-time research.

[6/24, 11:55 PM] OUR DAILY WISDOM: Meet Perplexity

That's exactly the problem Perplexity was built to solve. By the end of this lesson, you'll know what it does, how it works, and when to use it.

Before we go further, what is one common challenge in online research?

Search engines struggle with complex questions

There isn't enough information available online

There's a lot of information, but it's not synthesized into a clear answer

Correct answer

Right. The real challenge today usually isn't access to information — it's making sense of it. You often open multiple tabs, compare sources, and connect the dots yourself.

What Is Perplexity

completed check

Task completed

Your First Search

Ask Preplexity a question about the team's off-site.

Compare this to opening Google for the same question. What did Perplexity do that Google doesn't?

It searched more websites than Google would

It personalized results based on your location

It read multiple sources and gave one synthesized answer

Correct answer

Right. The volume of sources isn't the difference — Perplexity did the reading and condensing for you.

So Perplexity reads, synthesizes, and shows its work all in one step. But if that sounds familiar, you might be thinking of ChatGPT.

Perplexity vs. ChatGPT

Perplexity might look similar to ChatGPT at first — you type a question and get an answer.

You need to compare venue costs across three cities for next month's offsite. Which tool fits best?

ChatGPT — it can reason through the comparison based on what it knows

Google — it returns the most results to choose from

Perplexity — it searches the live web and cites every source

Correct answer

When you need current, verifiable information, Perplexity is the right call. Its web search is always on and every claim is cited. ChatGPT can search the web too, but it's built for conversation and creativity, not sourced research. Google returns links you'd have to open and compare yourself. Perplexity does that synthesis for you.

[6/24, 11:55 PM] OUR DAILY WISDOM: Correct answer

[6/24, 11:56 PM] OUR DAILY WISDOM: Perplexity's Difference

Google finds links.

ChatGPT generates responses, with optional web search.

Perplexity reads the live web and shows its work — every time.

[6/24, 11:57 PM] OUR DAILY WISDOM: Looking at the Interface

Now that you know what sets Perplexity apart, let's look at how it's laid out and how to find your way around it.

When you open Perplexity, you'll use two things consistently. New is where you tap to start a new thread. History is where your research history is stored so you can return to any search later.

[6/24, 11:57 PM] OUR DAILY WISDOM: Looking at the Interface

Now that you know what sets Perplexity apart, let's look at how it's laid out and how to find your way around it.

When you open Perplexity, you'll use two things consistently. New is where you tap to start a new thread. History is where your research history is stored so you can return to any search later.

[6/24, 11:58 PM] OUR DAILY WISDOM: When you run a search, you'll see the sources panel on the right. Every claim in the output links back to where it came from, so you can check any figure before using it.

[6/24, 11:59 PM] OUR DAILY WISDOM: You can click the menu at the top of any conversation to rename the thread, add a bookmark, or export it as a PDF, Markdown, or DOCX file.

You'll find all your saved threads in the left sidebar, ready to revisit whenever you need them.

[6/24, 11:59 PM] OUR DAILY WISDOM: Now that you know how Perplexity works, the next step is knowing when to reach for it.

[6/25, 12:00 AM] OUR DAILY WISDOM: Good Use or Bad Use?

Identify whether this is a good use of Perplexity.

Comparing four cities across cost, venues, and weather for an upcoming team offsite in Preplexity

Good use

Bad use

Amazing!

Exactly. Multiple variables, current data, sources to verify — that's where Perplexity earns its place. This is the kind of question that used to take 40 minutes of tab-switching.

[6/25, 12:02 AM] OUR DAILY WISDOM: What's the biggest way Perplexity changes the approach to research tasks?

It returns answers faster because it skips lower-quality websites

It pulls from more sources than a standard search engine would

It collapses searching, reading, comparing, and verifying into one step

Incorrect answer

Speed and source volume matter, but they don't change the workflow. The real difference is synthesis — Perplexity reads the sources and hands you a conclusion instead of a list of places to go look.

💡 When to Reach for Perplexity

Use Perplexity when your question involves multiple variables, needs current information, or requires comparing sources.

[6/25, 12:04 AM] OUR DAILY WISDOM: You're researching venues for a team offsite. Put these steps in the right order.

Drag the items below into order

Draggable item b was dropped over droppable area a

Amazing!

You're right on track with your approach

[6/25, 12:05 AM] OUR DAILY WISDOM: Key Takeaways

Perplexity is an AI-powered search engine — it reads the web and synthesizes the answer.

Unlike Google, it returns conclusions, not links. Unlike ChatGPT, every claim is cited by default.

Use it for multi-variable questions that need current, verifiable information.

[6/25, 12:05 AM] OUR DAILY WISDOM: That's the foundation of what Perplexity is and how it works. You've gone from blank search bar to a verified, saved research thread — and that was just the beginning.

[6/25, 12:07 AM] OUR DAILY WISDOM: What's Next?

Next, you'll learn how to mode for your task — because not every search needs the same approach, and knowing the difference is what separates a quick answer from a genuinely useful one. See you in the next lesson!

[6/25, 11:33 PM] OUR DAILY WISDOM: 11:34pm

Modes and Controls

You open Perplexity and start typing. The answers come back fine — usable, quick, good enough. Three searches later, you realize you've been running everything on default, and the results could have been sharper the whole time.

So, how do you make your searches deeper and more precise? In this lesson, you'll learn how to customize Perplexity with plans and modes to get exactly the results you need.

[6/25, 11:34 PM] OUR DAILY WISDOM: Modes and Controls

So, how do you make your searches deeper and more precise? In this lesson, you'll learn how to customize Perplexity with plans and modes to get exactly the results you need.

[6/25, 11:35 PM] OUR DAILY WISDOM: Scenario: Researching a Team Offsite

You're building the research case for a company off-site from the previous lesson. You need to:

Get a quick overview of city options

Compare costs and logistics in detail

Find credible data to justify the budget

Revisit earlier research as new questions come up

[6/25, 11:37 PM] OUR DAILY WISDOM: Modes and Controls

So, how do you make your searches deeper and more precise? In this lesson, you'll learn how to customize Perplexity with plans and modes to get exactly the results you need.

Scenario: Researching a Team Offsite

You're building the research case for a company off-site from the previous lesson. You need to:

Get a quick overview of city options

Compare costs and logistics in detail

Find credible data to justify the budget

Revisit earlier research as new questions come up

What's most likely to happen if you use the same default settings for every one of these searches?

Perplexity starts pulling from fewer sources over time

Each search takes longer as the questions get more complex

The answers become harder to read as the topic gets more detailed

Some searches return results that aren't deep or specific enough for the task

Incorrect answer

The issue isn't speed, volume, or readability, but how it's fitting. The same settings applied to very different questions will work well for some and fall short for others.

[6/25, 11:38 PM] OUR DAILY WISDOM: The same default settings applied to four very different tasks will fall short on at least some of them. That's what different plans solve. Let's go through it!

Plans — What's Available

Perplexity runs on three tiers:

The free plan gives you practically unlimited standard searches for research tasks.

Pro unlocks access to the latest AI models trained for higher accuracy for personal use.

Max goes further with unlimited usage and the most advanced AI reasoning models for heavy users

[6/25, 11:39 PM] OUR DAILY WISDOM: There are also two specialized plans for specific contexts.

Education Pro for students and educators: suited for academic research and complex reports at a steep discount.

Business for teams that need more than personal research: premium source citations from PitchBook and Statista, compliance certifications, SSO, and data privacy guarantees.

[6/25, 11:41 PM] OUR DAILY WISDOM: There are also two specialized plans for specific contexts.

Education Pro for students and educators: suited for academic research and complex reports at a steep discount.

Business for teams that need more than personal research: premium source citations from PitchBook and Statista, compliance certifications, SSO, and data privacy guarantees.

Plans control your access level. Modes control how Perplexity thinks about your query: how deeply it reasons, which models it uses, and how it structures its answer. The results are sourced, synthesized, and ready to use regardless of which mode you're in.

Best Mode

Best is the default mode — it picks the right model for your query automatically, whether that's a quick fact-check or a deep multi-source comparison.

On the free plan, Best runs silently in the background with no model selector visible. On Pro or Max, you'll see a Model dropdown where you can pick a specific model or stick with Best and let Perplexity decide.

[6/25, 11:42 PM] OUR DAILY WISDOM: Best Mode or Not?

Decide whether Best mode is the right fit for this situation.

A broad overview of city options for the offsite

Good fit

Bad fit

Amazing!

Best mode is exactly right here. You're orienting, not deciding — a fast, broad answer is what moves you forward at this stage.

[6/25, 11:42 PM] OUR DAILY WISDOM: Best mode gets you oriented fast. But some questions need more than a fast answer. They need thorough research across dozens of sources before anything is returned.

[6/25, 11:44 PM] OUR DAILY WISDOM: Your team lead asks: "Prague looks cheapest — but can you back that up with real data on costs, logistics, and what other companies have done for similar offsites?" What does that question need?

A faster, more detailed version of the cost breakdown already returned

A side-by-side summary of each city without a final recommendation

In-depth research across multiple sources with a structured, cited report

Correct answer

Exactly! This question needs evidence, not just an answer. That's exactly what Deep Research is built for. It searches broadly, synthesizes across sources, and returns something you can present.

[6/25, 11:46 PM] OUR DAILY WISDOM: Modes and Controls

So, how do you make your searches deeper and more precise? In this lesson, you'll learn how to customize Perplexity with plans and modes to get exactly the results you need.

Scenario: Researching a Team Offsite

You're building the research case for a company off-site from the previous lesson. You need to:

Get a quick overview of city options

Compare costs and logistics in detail

Find credible data to justify the budget

Revisit earlier research as new questions come up

What's most likely to happen if you use the same default settings for every one of these searches?

Perplexity starts pulling from fewer sources over time

Each search takes longer as the questions get more complex

The answers become harder to read as the topic gets more detailed

Some searches return results that aren't deep or specific enough for the task

Incorrect answer

The issue isn't speed, volume, or readability, but how it's fitting. The same settings applied to very different questions will work well for some and fall short for others.

The same default settings applied to four very different tasks will fall short on at least some of them. That's what different plans solve. Let's go through it!

Plans — What's Available

Perplexity runs on three tiers:

The free plan gives you practically unlimited standard searches for research tasks.

Pro unlocks access to the latest AI models trained for higher accuracy for personal use.

Max goes further with unlimited usage and the most advanced AI reasoning models for heavy users.

There are also two specialized plans for specific contexts.

Education Pro for students and educators: suited for academic research and complex reports at a steep discount.

Business for teams that need more than personal research: premium source citations from PitchBook and Statista, compliance certifications, SSO, and data privacy guarantees.

Best Mode

Best is the default mode — it picks the right model for your query automatically, whether that's a quick fact-check or a deep multi-source comparison.

completed check

Task completed

Best Mode or Not?

Decide whether Best mode is the right fit for this situation.

Best mode gets you oriented fast. But some questions need more than a fast answer. They need thorough research across dozens of sources before anything is returned.

Your team lead asks: "Prague looks cheapest — but can you back that up with real data on costs, logistics, and what other companies have done for similar offsites?" What does that question need?

A faster, more detailed version of the cost breakdown already returned

A side-by-side summary of each city without a final recommendation

In-depth research across multiple sources with a structured, cited report

Correct answer

Exactly! This question needs evidence, not just an answer. That's exactly what Deep Research is built for. It searches broadly, synthesizes across sources, and returns something you can present.

Deep Research

Deep Research doesn't just return an answer — it conducts research on your behalf. It breaks your question down, searches across dozens of sources, and synthesizes everything into a structured, cited report.

It's slower than Best mode, but it's built for questions where you need something you can actually stand behind — budget justifications, competitive analysis, anything where sourcing matters.

You now have two modes in your toolkit. The one you pick depends on where you are in the research — how much you already know, how complex the question is, and how much the answer needs to hold up to scrutiny.

[6/25, 11:46 PM] OUR DAILY WISDOM: Match the Mode to the Task

Match each mode to what it's built for.

Best mode

Deep Research

Fast, broad answers

Multi-source research with citations

Amazing!

You're right on track with your approach

[6/25, 11:48 PM] OUR DAILY WISDOM: Filtering by Source Type

So, you know that modes control how Perplexity thinks about your question. Source filters control where it looks, restricting results to a specific type before the search runs. All of them are available on the free plan, and you'll find them at the top of the screen when you start a new chat.

Filters narrow the search space for Perplexity. Here's what each one does:

Discover: Surfaces trending topics and news from across the web — good for staying current without a specific question in mind.

Finance: Pulls from official company financial disclosures and market data.

Health: Focused on medical sources, clinical data, and health publications.

Academic: Filters for scholarly papers and peer-reviewed journals.

Patents: Searches patent databases and intellectual property filings.

[6/25, 11:49 PM] OUR DAILY WISDOM: The best way to understand source filters is to see them in action. Imagine you need to check the US market before making any financial decisions for the trip. Instead of scrolling through news sites, you switch to Finance — and Perplexity pulls directly from market data and financial filings.

[6/25, 11:51 PM] OUR DAILY WISDOM: Perplexity Computer

Source filters narrow where Perplexity looks. But what if the task itself is too complex for a single search and require multiple steps, tools, and decisions along the way?

Perplexity Computer handles that. Think of it as an AI that doesn't just answer questions, but gets things done. You describe the task, and Computer works through it the same way you would, just without you doing any of the steps.

Back to the offsite. You need flight options from Paris to Prague — airlines, frequencies, price ranges, and a recommendation. Instead of opening three tabs and piecing it together yourself or chaining your requests, you hand it to Computer.

[6/25, 11:53 PM] OUR DAILY WISDOM: and financial filings.

Perplexity Computer

Source filters narrow where Perplexity looks. But what if the task itself is too complex for a single search and require multiple steps, tools, and decisions along the way?

Finding the Best Route

Use Computer to research flight options between your two shortlisted cities.

Finding the Best Route

Use Computer to research flight options between your two shortlisted cities.

Perplexity

Best flight connections from Paris to Prague — looking for frequency, duration, and average cost

Amazing!

You're right on track with your approach

[6/25, 11:55 PM] OUR DAILY WISDOM: What made Computer useful for this task compared to a standard search?

Select all possible answers

No manual cross-referencing needed — it handled the full task end-to-end

It searched more sources than usual and ranked them by reliability

It pulled from multiple sources and synthesized the result in one output

Almost right!

Computer works through the task the way you would — multiple sources and steps — and delivers one clean result. That's the difference between searching and doing.

That's a glimpse of what Computer can do. Later in this course, you'll go deeper — longer tasks, more complex workflows, and how to get the most out of it.

[6/25, 11:56 PM] OUR DAILY WISDOM: Key Takeaways

Free plan covers unlimited standard searches and all source filters

Pro and Max unlock unlimited Deep Research access when your volume or performance needs to grow.

Best mode is the default — automatic model selection in Pro and Max, handles most questions well.

Deep Research searches across dozens of sources and returns a structured, cited report.

Source filters narrow results by type — Academic, Finance, Health, and more — all on the free plan.

Perplexity Computer executes full workflows end-to-end — not just answers, but tasks done.

[6/25, 11:57 PM] OUR DAILY WISDOM: What's Next?

You now know how to get more precise answers — the right mode, the right sources, and a first look at what Computer can handle. Next, you'll go one level deeper: search operators that let you control exactly where Perplexity looks and what it brings back.

[6/27, 8:31 PM] OUR DAILY WISDOM: Search Operators & Filters

A well-structured prompt gets you closer to what you need — but not always all the way there. Sometimes the results come back from the wrong sources, in the wrong format, or from three years ago.

Search operators are how you close that gap: short commands that tell Perplexity exactly where to look, what type of content to surface, and how recent it needs to be.

[6/27, 8:32 PM] OUR DAILY WISDOM: What Are Search Operators

Search operators are short commands you add directly to your search. They control where Perplexity looks and what it surfaces. Think of them as filters you apply before the search runs.

[6/27, 8:33 PM] OUR DAILY WISDOM: Without operators, Perplexity decides which sources to pull from. With them, you decide. A personal question works fine without operators, but the moment source type, recency, or precision matters, operators close that gap.

[6/27, 8:41 PM] OUR DAILY WISDOM: Which type of search benefits most from using a search operator?

Looking up the best coffee shops in your neighborhood

Finding current pricing data for a competitive analysis

Correct answer

Operators earn their place when the source, format, or recency of results matters. A competitive analysis needs credible, current data in a usable format — that's exactly when operators add value.

[6/27, 8:42 PM] OUR DAILY WISDOM: Now imagine the same search principle applied to something where the source matters — like finding current pricing data or an industry report. For that, a standard search isn't enough.

That's where operators come in. There are four worth knowing, each one solving a different problem.

[6/27, 8:43 PM] OUR DAILY WISDOM: Here's what each operator does:

Site filters ("site:") restrict results to a specific domain.

Filetype filters ("filetype:") surface a specific document type, such as PDF or DOC.

Date filters ("after:"/"before:") force results within a specific time window.

Quotes match an exact phrase rather than individual words.

[6/27, 8:44 PM] OUR DAILY WISDOM: Restricting by Source

Let's start with "site:" — the operator that controls where Perplexity looks. Add it to any search, and Perplexity pulls only from that domain — no competing pages, no off-topic results. Use it when the source matters as much as the answer. See how that looks like on a simple personal question first.

[6/27, 8:48 PM] OUR DAILY WISDOM: Source-Restricted Search

You want to learn about home pruning techniques, but only from a trusted gardening institution.

Perplexity

How to prune roses site:rhs.org.uk

Amazing!

You're right on track with your approach

[6/27, 8:49 PM] OUR DAILY WISDOM: What did the "site:" operator do?

It signalled the question was technical, prompting Perplexity to prioritise more authoritative sources

It filtered the results after the search ran, removing lower-quality sources

It told Perplexity where to look before the search ran, so only that domain's content was eligible

Correct answer

"site:" is a pre-search filter — it sets the boundaries before Perplexity starts looking. Everything outside that domain is excluded from the start, which is why the results changed completely.

[6/27, 8:50 PM] OUR DAILY WISDOM: Search Operators & Filters

A well-structured prompt gets you closer to what you need — but not always all the way there. Sometimes the results come back from the wrong sources, in the wrong format, or from three years ago.

Search operators are how you close that gap: short commands that tell Perplexity exactly where to look, what type of content to surface, and how recent it needs to be.

What Are Search Operators

Search operators are short commands you add directly to your search. They control where Perplexity looks and what it surfaces. Think of them as filters you apply before the search runs.

Without operators, Perplexity decides which sources to pull from. With them, you decide. A personal question works fine without operators, but the moment source type, recency, or precision matters, operators close that gap.

Which type of search benefits most from using a search operator?

Looking up the best coffee shops in your neighborhood

Finding current pricing data for a competitive analysis

Correct answer

Operators earn their place when the source, format, or recency of results matters. A competitive analysis needs credible, current data in a usable format — that's exactly when operators add value.

Now imagine the same search principle applied to something where the source matters — like finding current pricing data or an industry report. For that, a standard search isn't enough.

That's where operators come in. There are four worth knowing, each one solving a different problem.

Here's what each operator does:

Site filters ("site:") restrict results to a specific domain.

Filetype filters ("filetype:") surface a specific document type, such as PDF or DOC.

Date filters ("after:"/"before:") force results within a specific time window.

Quotes match an exact phrase rather than individual words.

Restricting by Source

completed check

Task completed

Source-Restricted Search

You want to learn about home pruning techniques, but only from a trusted gardening institution.

What did the "site:" operator do?

It signalled the question was technical, prompting Perplexity to prioritise more authoritative sources

It filtered the results after the search ran, removing lower-quality sources

It told Perplexity where to look before the search ran, so only that domain's content was eligible

Correct answer

"site:" is a pre-search filter — it sets the boundaries before Perplexity starts looking. Everything outside that domain is excluded from the start, which is why the results changed completely.

That's how the source operator works! Note that it returned 10 sources only from the specific site you requested.

[6/27, 8:51 PM] OUR DAILY WISDOM: The same operator works just as well for professional research. Restricting to a trusted publication means credible, relevant sources appear from the start.

Now let's look at the next operator, which controls the type of document that comes back.

[6/27, 8:53 PM] OUR DAILY WISDOM: Finding Specific Document Types

Some information only exists inside documents — research reports, whitepapers, official papers — not on web pages. If what you need is in a PDF or a Word file rather than an article, web results won't get you there.

That's what "filetype:" solves. Add "filetype:pdf" or "filetype:doc" to any search, and Perplexity surfaces the documents directly — the reports and research papers that web pages only ever summarize. Let's say you need to find research on sleep patterns.

[6/27, 8:59 PM] OUR DAILY WISDOM: Finding Specific Document Types

Document-Type Search

Search for sleep patterns research restricted to PDF documents only.

Document-Type Search

Search for sleep patterns research restricted to PDF documents only.

Perplexity

Sleep patterns research filetype:pdf

Sleep patterns research, particularly from PDF studies, reveals key insights into how sleep cycles, durations, and disruptions affect health, cognition, and daily performance.

Core Sleep Architecture

Human sleep cycles through non-REM (N1-N3) and REM stages roughly 4-6 times nightly, with each cycle lasting 90-110 minutes (1955: Aserinsky/Kleitman REM discovery); early cycles emphasize deep N3 slow-wave sleep, while later ones increase REM for memory processing.

Research Priorities

It prioritizes biomarkers for sleep quality, chronotherapeutics for chronic conditions, and tools for early detection in youth to prevent lifelong issues (1997 NHLBI influences).

Population studies highlight 30-40% of U.S. adults facing sleep deficiency, costing $411 billion yearly in health and productivity losses.

Source: Sleepresearchsociety.com, Sciencedirect.com

Amazing!

You're right on track with your approach

[6/27, 9 PM] OUR DAILY WISDOM: filetype:pdf" went straight to the source documents where the actual data lives. Perplexity also lets you download research papers from the search results.

But sometimes Perplexity returns data more than 10 years old, and in a fast-moving research field, that's a problem. When the data needs to be current, you need to control the time window. That's what the next operator is for.

[6/27, 9:00 PM] OUR DAILY WISDOM: Filtering by Date

Date filters control exactly which time range your results come from. Add "after:" or "before:" to any search, and Perplexity only returns results inside that window — everything outside it is excluded before the search runs.

Use them whenever an old source would actively mislead rather than inform: pricing data, market reports, product releases, regulatory updates. Let's apply one now.

[6/27, 9:02 PM] OUR DAILY WISDOM: Filtering by Date

Use them whenever an old source would actively mislead rather than inform: pricing data, market reports, product releases, regulatory updates. Let's apply one now.

Date Filter

Search for sleep patterns research — this time filtered to the last two years only.

Date Filter

Search for sleep patterns research — this time filtered to the last two years only.

Perplexity

Sleep patterns research after:2025-01-01

Recent sleep patterns research since January 2025 has advanced understanding of chronotypes, trends in disturbances, and health links using large datasets.

Chronotype Subtypes

A February 2026 McGill study in Nature Communications identified five biological chronotype subtypes within night owls (3) and early birds (2), using AI on 27K+ UK Biobank adults' brain imaging, questionnaires, and records; subtypes link to unique risks like depression, cardiovascular issues, or cognitive strengths beyond simple bedtimes. These reflect genetic-lifestyle interactions, urging personalized sleep support over binary labels.

Rising Sleep Issues

A September 2025 Danish study (Science Advances) tracked 15-45-year-olds (2010-2021), finding self-reported sleep problems rose from 34% to 49%, with melatonin prescriptions surging 10-fold; machine learning on 2.2M life-course records clustered patterns tying sleep to pain, mental health, or early adversity.

SLEEP 2025 featured Miller School posters on sleep's Alzheimer's role via Fitbit-tracked stages and biomarkers, plus discrimination-insomnia links are stronger in females.

Sources: McGill.com, ScienceDirect.com, SleepResearchSociety.org

Amazing!

You're right on track with your approach

[6/27, 9:07 PM] OUR DAILY WISDOM: Filtering by Date

Use them whenever an old source would actively mislead rather than inform: pricing data, market reports, product releases, regulatory updates. Let's apply one now.

Date Filter

Search for sleep patterns research — this time filtered to the last two years only.

Date Filter

Search for sleep patterns research — this time filtered to the last two years only.

Perplexity

Sleep patterns research after:2025-01-01

Recent sleep patterns research since January 2025 has advanced understanding of chronotypes, trends in disturbances, and health links using large datasets.

Chronotype Subtypes

Rising Sleep Issues

SLEEP 2025 featured Miller School posters on sleep's Alzheimer's role via Fitbit-tracked stages and biomarkers, plus discrimination-insomnia links are stronger in females.

Sources: McGill.com, ScienceDirect.com, SleepResearchSociety.org

Amazing!

You're right on track with your approach

[6/27, 9:08 PM] OUR DAILY WISDOM: What does the date filter do to your search results?

It sets a hard boundary — everything outside that window was excluded before the search ran

It tells Perplexity to prioritize more recent sources while still including older ones lower down

It improves result quality by signalling the search needs current information

Correct answer

Date filter is a pre-search boundary, not a ranking signal. Everything outside a specific window was excluded entirely, which is why only that product cycle appeared, nothing before or after.

[6/27, 9:10 PM] OUR DAILY WISDOM: Every result inside the window, nothing outside it. That's the date filter keeping the results where they need to be.

Now there's one more precision tool worth knowing — for when the words are right, but the context needs to be more exact.

Exact Phrase Matching

Putting a phrase in quotes tells Perplexity to match those words exactly in that order. With quotes, a search for "machine learning algorithm list" returns only results where that exact phrase appears. Let's say you need to find Apple's financial reports specifically. That's the kind of search where quotes make the difference.

[6/27, 9:11 PM] OUR DAILY WISDOM: Now there's one more precision tool worth knowing — for when the words are right, but the context needs to be more exact.

Exact Phrase Matching

Exact Phrase

Search for a specific Apple concept using exact phrase matching — without quotes, the individual words return unrelated results.

Exact Phrase

Search for a specific Apple concept using exact phrase matching — without quotes, the individual words return unrelated results.

Perplexity

"Apple's financial report" for 2026

Apple's published financial results for 2026 so far cover only the first fiscal quarter (Q1 FY26)

Segment and operational highlights

— iPhone: Revenue around $85.3 billion, up roughly 23% year‑over‑year, marking a record‑setting quarter.

— Services: Revenue about $30 billion, up 14% year‑over‑year, also a record.

— Mac and Wearables: Mac revenue declined slightly (about ‑7%), while Wearables, Home, and Accessories were down about ‑2%, mainly due to product‑mix and supply constraints.

Source: Apple.com

Amazing!

You're right on track with your approach

[6/27, 9:16 PM] OUR DAILY WISDOM: See how "Financial Reports" came back specific — not scattered results about Apple's earnings calls, or investor commentary in unrelated contexts. The quotes matched the exact phrase before the search ran, which is why every result pointed to the same specific thing.

[6/27, 9:17 PM] OUR DAILY WISDOM: Combining Operators

Each operator solves one problem, but combined, they close all the gaps at once — source, format, recency, and precision in a single search. Let's add everything together.

[6/27, 9:18 PM] OUR DAILY WISDOM: Apple’s financial reports"

site:apple.com

filetype:pdf

after:2026-01-01

[6/27, 9:21 PM] OUR DAILY WISDOM: Combining Operators

Each operator solves one problem, but combined, they close all the gaps at once — source, format, recency, and precision in a single search. Let's add everything together.

The Full Operator Search

Run the Apple search with all four operators combined — exact phrase, site filter, filetype, and date filter.

The Full Operator Search

Run the Apple search with all four operators combined — exact phrase, site filter, filetype, and date filter.

Perplexity

"Apple’s financial reports", site:apple.com, filetype:pdf, after:2026-01-01

Apple released its FY2026 Q1 Consolidated Financial Statements PDF on apple.com/newsroom after January 1, 2026, covering results through December 27, 2025.

Key Financial Highlights

Revenue: $143.8 billion (up 16% YoY), with iPhone and Services hitting all-time highs.

EPS: $2.84 (up 19% YoY).

Operating Cash Flow: Nearly $54 billion, enabling $32 billion returned to shareholders.

This PDF includes detailed statements of operations, cash flows, and balance sheets, accessible via Apple's investor relations section. No other major "Financial Reports" developments (e.g., Q2) appear post-January in PDF format yet.

Source: Apple.com

Amazing!

You're right on track with your approach

[6/27, 9:22 PM] OUR DAILY WISDOM: Search Operators & Filters

A well-structured prompt gets you closer to what you need — but not always all the way there. Sometimes the results come back from the wrong sources, in the wrong format, or from three years ago.

Search operators are how you close that gap: short commands that tell Perplexity exactly where to look, what type of content to surface, and how recent it needs to be.

What Are Search Operators

Search operators are short commands you add directly to your search. They control where Perplexity looks and what it surfaces. Think of them as filters you apply before the search runs.

Which type of search benefits most from using a search operator?

Looking up the best coffee shops in your neighborhood

Finding current pricing data for a competitive analysis

Correct answer

Operators earn their place when the source, format, or recency of results matters. A competitive analysis needs credible, current data in a usable format — that's exactly when operators add value.

Now imagine the same search principle applied to something where the source matters — like finding current pricing data or an industry report. For that, a standard search isn't enough.

That's where operators come in. There are four worth knowing, each one solving a different problem.

Here's what each operator does:

Site filters ("site:") restrict results to a specific domain.

Filetype filters ("filetype:") surface a specific document type, such as PDF or DOC.

Date filters ("after:"/"before:") force results within a specific time window.

Quotes match an exact phrase rather than individual words.

Restricting by Source

completed check

Task completed

Source-Restricted Search

You want to learn about home pruning techniques, but only from a trusted gardening institution.

What did the "site:" operator do?

It signalled the question was technical, prompting Perplexity to prioritise more authoritative sources

It filtered the results after the search ran, removing lower-quality sources

It told Perplexity where to look before the search ran, so only that domain's content was eligible

Correct answer

"site:" is a pre-search filter — it sets the boundaries before Perplexity starts looking. Everything outside that domain is excluded from the start, which is why the results changed completely.

That's how the source operator works! Note that it returned 10 sources only from the specific site you requested.

The same operator works just as well for professional research. Restricting to a trusted publication means credible, relevant sources appear from the start.

Now let's look at the next operator, which controls the type of document that comes back.

Finding Specific Document Types

completed check

Task completed

Document-Type Search

Search for sleep patterns research restricted to PDF documents only.

"filetype:pdf" went straight to the source documents where the actual data lives. Perplexity also lets you download research papers from the search results.

Filtering by Date

Use them whenever an old source would actively mislead rather than inform: pricing data, market reports, product releases, regulatory updates. Let's apply one now.

completed check

Task completed

Date Filter

Search for sleep patterns research — this time filtered to the last two years only.

What does the date filter do to your search results?

It sets a hard boundary — everything outside that window was excluded before the search ran

It tells Perplexity to prioritize more recent sources while still including older ones lower down

It improves result quality by signalling the search needs current information

Correct answer

Date filter is a pre-search boundary, not a ranking signal. Everything outside a specific window was excluded entirely, which is why only that product cycle appeared, nothing before or after.

Every result inside the window, nothing outside it. That's the date filter keeping the results where they need to be.

Now there's one more precision tool worth knowing — for when the words are right, but the context needs to be more exact.

Exact Phrase Matching

completed check

Task completed

Exact Phrase

Search for a specific Apple concept using exact phrase matching — without quotes, the individual words return unrelated results.

See how "Financial Reports" came back specific — not scattered results about Apple's earnings calls, or investor commentary in unrelated contexts. The quotes matched the exact phrase before the search ran, which is why every result pointed to the same specific thing.

Combining Operators

Each operator solves one problem, but combined, they close all the gaps at once — source, format, recency, and precision in a single search. Let's add everything together.

completed check

Task completed

The Full Operator Search

Run the Apple search with all four operators combined — exact phrase, site filter, filetype, and date filter.

What did combining all four operators give you that a standard search wouldn't?

Control over the source, format, and recency before the search ran, so the first result was already the right one

A result that required less verification because the operators pre-validated the sources

A more comprehensive answer, because each operator added a new layer of research depth

Incorrect answer

Operators don't add research depth or validate sources — they filter which results are eligible before Perplexity ranks them. Comprehensive answers come from the quality of the question, not the number of operators. And verification is still your responsibility — operators narrow the pool, they don't check what's in it.

[6/27, 9:25 PM] OUR DAILY WISDOM: Match the Need to the Operator

Match each research need to the operator that solves it.

You need results only from a specific trusted domain

You need an exact metric rather than scattered mentions of the words

You need results published in the last 12 months only

You need a downloadable report rather than a web page

"site:"

Quotes

Date operator

"file:"

Amazing!

You're right on track with your approach

[6/27, 9:25 PM] OUR DAILY WISDOM: Key Takeaways

Site filters ("site:") restrict results to a specific domain: Use when the source matters as much as the answer.

Filetype filters ("filetype:") surface reports and research documents instead of web page summaries.

Date filters ("after:"/"before:") force recency: Essential for pricing data, market reports, and anything time-sensitive.

Quotes match an exact phrase rather than individual words: Use for specific metrics, terms, or titles.

Combined, all four give you a precise query: Source, format, recency, and phrase work together.

[6/27, 9:28 PM] OUR DAILY WISDOM: You now know how to tell Perplexity not just what to find, but where to look, what format to return, and how recent the results need to be.

Next, you'll learn how to organize everything into one persistent workspace, so your research builds on itself, your files are always at hand, and your whole team works from the same place. See you there!

[6/30, 11:26 PM] OUR DAILY WISDOM: Your Research HQ

Two weeks into planning the offsite. Notes in three apps, tabs you're scared to close, and a Perplexity history you re-explain every session. There's a better way to keep it all in one place — and it's called Spaces.

By the end of this lesson, you'll know how to set up a Space, upload your own files, and query directly against them — so your research builds on itself.

[6/30, 11:27 PM] OUR DAILY WISDOM: What Is a Space?

A Space is a persistent project environment inside Perplexity. It keeps your research threads, uploaded files, custom instructions, and collaborators in one place, so every session picks up where the last one left off.

[6/30, 11:28 PM] OUR DAILY WISDOM: Without a Space, every new thread starts fresh. You carry the overhead of re-establishing context, re-uploading files, and re-explaining priorities every time. Let's try creating a Space.

Creating Your First Space

Click New Space in the right sidebar, give it a name like "Team Offsite 2026," and you're in. From here, you can add instructions, upload files, and invite collaborators.

[6/30, 11:29 PM] OUR DAILY WISDOM: Before you create your first Space, Perplexity shows you a selection of pre-built templates — things like Trip Planner Pro, Research Assistant, and more. Pick one that fits your task and it opens ready to use.

For your offsite task, you can use the "Trip Planner Pro" template.

[6/30, 11:31 PM] OUR DAILY WISDOM: Templates are designed for quick, one-off use — they're not saved to your Spaces by default, and they work best as-is rather than something you'd customize or share. For a project you'll keep coming back to, there's a better move.

💡 Getting the Most Out of Templates

Templates are a starting point, not a finished workspace. To unlock the full power of Spaces — sharing, file uploads, custom instructions — copy the template's instructions and paste them into your own new Space. Everything then works the way you'd expect.

That's exactly what we'll do with the Trip Planner Pro template. Create a new Space called "Team Offsite 2026," and paste the instructions under Space instructions. From that point, every search in the Space runs with that framing already built in.

[6/30, 11:34 PM] OUR DAILY WISDOM: The custom instruction can be as detailed as you need it to be. You can set the tone, format, priorities, and constraints.

In this case, it's set up as a travel planning assistant that asks clarifying questions, builds day-by-day itineraries, and includes local tips. Everything typed here runs silently behind every search in the Space. Let's run a search with that instruction.

[6/30, 11:35 PM] OUR DAILY WISDOM: The custom instruction can be as detailed as you need it to be. You can set the tone, format, priorities, and constraints.

Setting Up the Space

Open the "Trip Planner Pro" template, and type a prompt to plan an offsite.

Setting Up the Space

Open the "Trip Planner Pro" template, and type a prompt to plan an offsite.

Perplexity

Help me plan a team offsite for 10 people to Prague for 3 days.

Great choice — Prague is perfect for a 3-day cross-team alignment offsite! Its mix of historic charm, walkable center, and vibrant energy makes it ideal for bonding while tackling strategy sessions. Here's a tailored, realistic itinerary for your group of 10 that blends focused work, light team-building, and downtime.

Logistics: Central hotel (e.g., Four Seasons): $200-300/night. Total budget: $800-1200/person (excl. flights). Use Bolt/Uber + 72h tram pass ($15).

Day 1: Arrival & Kickoff

— Arrive, lunch at Lokál ($15).

— 3-7 p.m.: Alignment workshop + Old Town treasure hunt ($50/pp).

— Dinner at U Fleků beer hall ($30).

Day 2: Deep Work

— 9 a.m.-1 p.m.: Strategy sessions (roadmap/SWOT).

— 2-5 p.m.: Axe throwing/escape room ($30/pp).

— Evening: Prague Castle stroll + dinner at Eska ($40).

Day 3: Wrap-Up

— 9 a.m.-Noon: Review actions + feedback.

— 1-3 p.m.: Cooking class or Vltava cruise ($30-50).

— Depart.

Sources: Triphunter.com, Prague.eu

Amazing!

You're right on track with your approach

[6/30, 11:36 PM] OUR DAILY WISDOM: See how the template already knew how to structure around budget, logistics, and local tips. That's the custom instruction doing its job — your framing was built in before you even asked the first question.

Notice the follow-up suggestions at the bottom of the answer. Perplexity generates these automatically — a quick way to keep the research moving without having to think about what to ask next.

[6/30, 11:37 PM] OUR DAILY WISDOM: Asking Questions About Your Own Files

Perplexity also lets you upload files and search within them. When you run a thread inside a Space with uploaded files, it pulls answers from those files, not from the web. This means that your documents become the source.

To see how that works, you first need a file. Let's ask Perplexity to estimate the off-site budget, then turn that into a file you can upload.

[6/30, 11:40 PM] OUR DAILY WISDOM: Creating the Budget Estimate

Ask Perplexity to generate an estimated budget breakdown for the offsite that you can save and upload.

Perplexity

Create an estimated budget breakdown for a 10-person team offsite in Prague

Amazing!

You're right on track with your approach

[6/30, 11:41 PM] OUR DAILY WISDOM: Great! You have a file that Perplexity can read and answer questions about, using your actual numbers, not general web data. Let's put that to use.

[6/30, 11:42 PM] OUR DAILY WISDOM: Creating the Budget Estimate

Ask Perplexity to generate an estimated budget breakdown for the offsite that you can save and upload.

Great! You have a file that Perplexity can read and answer questions about, using your actual numbers, not general web data. Let's put that to use.

Asking About Your Budget

With the budget uploaded to your Space, ask Perplexity which categories are at the highest risk of going over.

Asking About Your Budget

With the budget uploaded to your Space, ask Perplexity which categories are at the highest risk of going over.

Perplexity

Based on the *uploaded file*, identify which categories are at the highest risk

From that budget table, the categories at highest risk of overruns are:

- Flights / travel — Marked as "depends on where you fly from," which is highly variable and sensitive to booking time, route availability, and airline choice. A flat 400 EUR per person can easily be exceeded if you're not flying from a nearby hub or if you book late.

- Accommodation — "Mid‑range hotel, 2 nights" at 360 EUR per person assumes good group rates and average seasonality; prices can spike for popular dates, trade fairs, or if you need very central locations, so this line can move a lot.

- Activities — "Team building + light extras" at 150 EUR per person is quite assumption‑heavy; costs will swing depending on whether you choose premium guided experiences versus simpler self‑guided or low‑cost options.

- Food & drinks — 200 EUR per person for all breakfasts, lunches, and two dinners is okay if you stay mid‑range, but nicer restaurants, extra drinks, or adding a special "wow" dinner can push this up.

Local transport, meeting space & AV, and the 15% misc/buffer are relatively less risky because they’re easier to fix‑price ahead of time (hotel meeting packages, pre‑booked transfers) and the buffer is already meant to absorb small fluctuations.

Amazing!

You're right on track with your approach

[6/30, 11:43 PM] OUR DAILY WISDOM: Now you can see which categories carry the most risk. The next question is what to do about them — specifically, where the budget can be brought down without changing the offsite team's spirit. Let's ask Perplexity.

[6/30, 11:44 PM] OUR DAILY WISDOM: completed check

Task completed

Asking About Your Budget

With the budget uploaded to your Space, ask Perplexity which categories are at the highest risk of going over.

Now you can see which categories carry the most risk. The next question is what to do about them — specifically, where the budget can be brought down without changing the offsite team's spirit. Let's ask Perplexity.

Finding Where to Cut

Ask Perplexity for specific ways to reduce costs in the highest-risk categories from your budget.

Finding Where to Cut

Ask Perplexity for specific ways to reduce costs in the highest-risk categories from your budget.

Perplexity

Based on the offsite budget estimate, what are specific savings options for flights and accommodation without compromising the group experience?

Flights

- Fly Tue-Thu or at less popular times.

- Compare fares on aggregators, then request a 10-person group booking.

- Use direct low-cost routes where practical.

- Limit add-ons: cabin bag + personal item, no paid seat selection unless needed.

- Book 8-10 weeks ahead and avoid expensive fully flexible fares.

Accommodation

- Choose clean, business-friendly 3-4* hotels, not boutique luxury.

- Stay 5-10 minutes outside the absolute center to lower rates.

- Ask for a package: rooms + meeting space + AV + coffee breaks.

- Pick rates with breakfast included.

- Avoid peak dates and major event periods.

Don't cut

Avoid long layovers, overnight travel, poorly reviewed hotels, or inconvenient locations. These savings usually hurt the whole offsite.

Sources: Tripmakery, Expedia, Skyscanner, Kayak.co

Amazing!

You're right on track with your approach

[6/30, 11:45 PM] OUR DAILY WISDOM: Specific platforms, price comparisons, and booking strategies — all cited so you can verify before acting. Sources pointing you toward cheaper alternatives you wouldn't have found in a general search. And now your whole team can benefit from it, not just you. Let's try bringing them in.

Sharing a Space

Once your Space is set up, click Share to share it and add collaborators. They join with access to the same threads, files, and instructions you've already built.

[6/30, 11:47 PM] OUR DAILY WISDOM: Good or Bad Use?

Decide whether this is a good use of Spaces or not.

Inviting your coordinator to join the existing "Team Offsite 2026" Space.

Good use

Bad use

Amazing!

Inviting a collaborator into an existing Space means they start with full context — the files, the instructions, the threads. They contribute forward rather than duplicating work that's already done.

[6/30, 11:47 PM] OUR DAILY WISDOM: And just like that, your coordinator is working from the same files, the same instructions, and the same research you've already built. That's Spaces doing what they're designed for.

[6/30, 11:48 PM] OUR DAILY WISDOM: There are pre-built templates for other situations as well.

If you're working on a piece of writing, the "Writing Coach" template gives you clear, actionable feedback on tone, structure, and clarity — just paste any text you want to improve.

If you're preparing for a job interview, the "Interview Roleplay Coach" template runs practice sessions and gives feedback based on your answers.

Now, let's recap the full process from start to finish.

[6/30, 11:51 PM] OUR DAILY WISDOM: situations as well.

If you're working on a piece of writing, the "Writing Coach" template gives you clear, actionable feedback on tone, structure, and clarity — just paste any text you want to improve.

If you're preparing for a job interview, the "Interview Roleplay Coach" template runs practice sessions and gives feedback based on your answers.

Now, let's recap the full process from start to finish.

Building a Space From Scratch

You're setting up the Team Offsite 2026 Space. Put these steps in the right order.

Building a Space From Scratch

You're setting up the Team Offsite 2026 Space. Put these steps in the right order.

Drag the items below into order

Draggable item b was dropped over droppable area a

Almost right!

You're close. Give it another try!

[6/30, 11:52 PM] OUR DAILY WISDOM: Key Takeaways

Spaces keep threads, files, instructions, and collaborators in one persistent place.

Templates give you a pre-configured starting point — copy the instructions into your own Space to unlock sharing, file uploads, and custom editing.

Custom instructions frame every search automatically.

Uploaded files become sources you can ask questions about directly.

Collaborators join with full context — no re-explaining needed.

[6/30, 11:53 PM] OUR DAILY WISDOM: What's Next?

You now know how to build a research environment that holds context, works with your own files, and brings your team in without starting over.

In the next unit, you'll go deeper into Perplexity Computer: how to hand it full workflows, not just single tasks, and what it can handle that no search mode can. See you there!

[7/2, 9:40 PM] OUR DAILY WISDOM: Turn Research Into a Workflow

You've compared cities, built a Space, flagged budget risks, and mapped out cost-saving options. The research is solid. But now you need the next crucial step — a finished plan. That's a different kind of task. And Perplexity has a different tool for it.

[7/2, 9:41 PM] OUR DAILY WISDOM: Perplexity Computer

Computer is a separate workspace inside Perplexity. You'll find it in the left sidebar. It's not a search mode or a research feature. It's where you hand off tasks that have multiple steps and need a finished result at the end.

[7/2, 9:42 PM] OUR DAILY WISDOM: Computer works through a task the way you would — it searches, reads pages, cross-references sources, uses your files, and keeps going until the deliverable is ready. The difference is it does all of that without you driving every step.

[7/2, 9:44 PM] OUR DAILY WISDOM: What makes Computer different from running a Deep Research query in Perplexity?

Computer executes a multi-step task and produces not just a report but a finished deliverable

Computer works faster because it runs multiple searches at the same time

Computer searches more sources and returns a longer, more detailed report

Computer uses more advanced AI models than Deep Research

Incorrect answer

The difference isn't speed, length, or model quality. It's actually what gets handed back. Deep Research returns research, and Computer returns a deliverable

[7/2, 9:45 PM] OUR DAILY WISDOM: What Computer Is For

Not every task needs Computer. The right signal is this: does the task have multiple steps and a specific output at the end?

Here's how Computer fits alongside everything else:

Search answers a single question

Deep Research produces a sourced research report

Spaces keep context, files, and threads in one place

Computer executes a multi-step task and delivers the finished result

[7/2, 9:55 PM] OUR DAILY WISDOM: Computer or Not?

Decide whether this is a good use of Computer.

"What's the average hotel price in Prague in July?"

Good use

Too small

Incorrect

Computer is built for multi-step workflows. Sending a single factual question to Computer is like hiring a project manager to check one price. The tool is doing far less than it's built for.

[7/2, 9:56 PM] OUR DAILY WISDOM: Computer or Not?

Decide whether this is a good use of Computer.

"What's the average hotel price in Prague in July?"

Good use

Too small

Amazing!

Right on! A single lookup doesn't need Computer. A quick search returns this in seconds. Computer is for tasks with multiple steps and a real deliverable at the end.

[7/2, 9:57 PM] OUR DAILY WISDOM: The clearest way to see Computer in action is through the offsite task that's already in progress.

Starting a Task in Computer

Click Computer in the sidebar, and you'll see a text input field. This is where you describe what you need. Remember that the task shouldn't be a search query, but a clear outcome with context.

[7/2, 9:59 PM] OUR DAILY WISDOM: The clearest way to see Computer in action is through the offsite task that's already in progress.

Starting a Task in Computer

Click Computer in the sidebar, and you'll see a text input field. This is where you describe what you need. Remember that the task shouldn't be a search query, but a clear outcome with context.

[7/2, 10:11 PM] OUR DAILY WISDOM: Hand Off the Workflow

Ask Computer to turn your offsite research into a planning brief.

Perplexity

Use

the

Team

Offsite

2026

Space

create

[deliverable type]

for

[event type]

Include

[planning elements]

and

flag

[review focus]

assumptions or missing information

the Prague team offsite

schedule, budget risks, logistics, and next steps

a planning brief

[7/2, 10:14 PM] OUR DAILY WISDOM: The clearest way to see Computer in action is through the offsite task that's already in progress.

Starting a Task in Computer

Click Computer in the sidebar, and you'll see a text input field. This is where you describe what you need. Remember that the task shouldn't be a search query, but a clear outcome with context.

Before you type, think through four things:

What you need produced

What context Computer should use

What the constraints are

What the output should look like

The clearer the handoff, the better the result is suited for you.

Hand Off the Workflow

Ask Computer to turn your offsite research into a planning brief.

Hand Off the Workflow

Ask Computer to turn your offsite research into a planning brief.

Perplexity

Use the Team Offsite 2026 Space to create a planning brief for the Prague team offsite. Include schedule, budget risks, logistics, and next steps and flag assumptions or missing information.

Prague Team Offsite Planning Brief

Executive summary

The Team Offsite 2026 Space supports a 10-person Prague offsite with a 3-day working concept and one confirmed event date: Wednesday, August 19, 2026. The strongest baseline is a cross-team alignment retreat combining strategy work, team bonding, Prague cultural activities, and practical action planning. The current plan is workable, but several core decisions remain open: whether August 19 is the single event day or the anchor date within a 3-day trip, the exact venue, final schedule, attendee list, flight origins, rooming policy, and food/dietary requirements.

Budget guidance in the Space has two useful frames. The saved Prague reference itinerary uses USD 800-1200 per person excluding flights, assuming a central hotel around a higher-comfort tier, Bolt/Uber plus 72-hour tram passes, and a mix of paid activities and group meals. An earlier detailed budget model estimated EUR 1000-1600 per person including flights, with a sample total of roughly EUR 13,900 for 10 people, rounded to EUR 14,000 for safety. The main cost risks are flights, accommodation, activities, and food upgrades, while meeting space, AV, and local transport should be easier to control if fixed early.

Amazing!

You're right on track with your approach

[7/2, 10:15 PM] OUR DAILY WISDOM: Notice what came back — a structured document: executive summary, confirmed facts pulled from your Space, open assumptions flagged for review. Computer shows its work and asks what to do next.

[7/2, 10:25 PM] OUR DAILY WISDOM: Workflows: Ready-Made Tasks

Computer also has pre-built Workflows — guided task templates for common deliverables. You'll find them under Customise in the sidebar.

[7/2, 10:27 PM] OUR DAILY WISDOM: Workflows cover a wide range of use cases: marketing, creative, career, productivity, sales, and many more. Pick one that fits, fill in your specifics, and Computer runs the whole process for you.

[7/2, 10:28 PM] OUR DAILY WISDOM: For offsite, Event prep template fits the best. With it, you can create an event brief, build and deploy a landing page, set up calendar invites, and manage RSVPs. Fill in the information and let Computer do the rest.

[7/2, 10:30 PM] OUR DAILY WISDOM: Describe the Task for Computer

Give Computer the key details it needs to create a planning brief for your offsite.

Perplexity

A 10-person 3-day team offsite in Prague. The confirmed date is August 19, 2026. Flight origins, rooming policy, and dietary requirements are TBC.

I'll set up the offsite package around the confirmed Prague start date, keeping unknowns like flight origins, rooming, and dietary needs explicitly marked as TBC. I'm going to prepare the brief and email templates, delegate the landing page build, and check whether Google Calendar is available before attempting any calendar setup.

Amazing!

You're right on track with your approach

[7/2, 10:32 PM] OUR DAILY WISDOM: Here's what came back. Computer built a full event package: a deployed landing page plus a structured event document you can view, edit, download, or publish directly. For the calendar invite, it flagged that authorization is still pending and left a ready-to-send draft for when you confirm the timing.

[7/2, 10:40 PM] OUR DAILY WISDOM: That's a lot done in one task. But before you share any of it, it's worth checking what Computer assumed along the way.

[7/2, 10:42 PM] OUR DAILY WISDOM: That's a lot done in one task. But before you share any of it, it's worth checking what Computer assumed along the way.

You review the Event prep output and notice two things: the landing page lists "venue TBC" and the agenda assumes all 10 people arrive by noon on August 19. What do you do before sharing?

Select all possible answers

Ask Computer to mark the venue as an open decision and explain the impact on the agenda

Ask Computer to add a note that arrival times need confirming before the schedule is finalised

Share it as-is — "TBC" signals to the team that details are still open

Ask Computer to flag which sections will need updating once the venue and arrivals are confirmed

Incorrect answer

"TBC" without context leaves the reader guessing what to do next. Computer can mark assumptions clearly and flag their downstream impact — that's a one-line follow-up request, not a full revision.

[7/2, 10:44 PM] OUR DAILY WISDOM: That's a lot done in one task. But before you share any of it, it's worth checking what Computer assumed along the way.

You review the Event prep output and notice two things: the landing page lists "venue TBC" and the agenda assumes all 10 people arrive by noon on August 19. What do you do before sharing?

Select all possible answers

Ask Computer to mark the venue as an open decision and explain the impact on the agenda

Ask Computer to add a note that arrival times need confirming before the schedule is finalised

Share it as-is — "TBC" signals to the team that details are still open

Ask Computer to flag which sections will need updating once the venue and arrivals are confirmed

Incorrect answer

"TBC" without context leaves the reader guessing what to do next. Computer can mark assumptions clearly and flag their downstream impact — that's a one-line follow-up request, not a full revision.

Now the brief is clean and the assumptions are visible. One more thing before you share — the constraints just changed.

[7/2, 10:46 PM] OUR DAILY WISDOM: That's a lot done in one task. But before you share any of it, it's worth checking what Computer assumed along the way.

You review the Event prep output and notice two things: the landing page lists "venue TBC" and the agenda assumes all 10 people arrive by noon on August 19. What do you do before sharing?

Select all possible answers

Ask Computer to mark the venue as an open decision and explain the impact on the agenda

Ask Computer to add a note that arrival times need confirming before the schedule is finalised

Share it as-is — "TBC" signals to the team that details are still open

Ask Computer to flag which sections will need updating once the venue and arrivals are confirmed

Incorrect answer

"TBC" without context leaves the reader guessing what to do next. Computer can mark assumptions clearly and flag their downstream impact — that's a one-line follow-up request, not a full revision.

Now the brief is clean and the assumptions are visible. One more thing before you share — the constraints just changed.

Your team lead updated the brief: the budget is 15% lower and one person will arrive late on Day 1. The main team-building activity stays. Everything else needs to adjust.

[7/2, 10:48 PM] OUR DAILY WISDOM: Revise the Plan

Ask Computer to update the offsite plan based on the new constraints.

Perplexity

Revise

the

offsite

plan

because

the budget is 15% lower

and

one

person

arrives

late

Day

Keep

[priority]

and

update

[sections to revise]

Almost right

Almost there! Review the steps and try again

[7/2, 10:54 PM] OUR DAILY WISDOM: Revise the Plan

Ask Computer to update the offsite plan based on the new constraints.

Perplexity

Revise the offsite plan because the budget is 15% lower and one person arrives late on Day 1. Keep the main team-building activity and update schedule, accommodation, and budget notes.

I'll revise the package around two concrete changes: a 15% budget reduction and one attendee joining late on Day 1, while preserving the core team-building activity. I'll update the brief document first, then adjust the landing page schedule/logistics and redeploy it.

Amazing!

You're right on track with your approach

[7/2, 10:55 PM] OUR DAILY WISDOM: Notice what happens before the answer arrives — Computer runs tasks in parallel, thinks through dependencies, reads the existing files, and explains its approach before making any changes. This is reasoning through what the budget cut affects.

[7/2, 10:57 PM] OUR DAILY WISDOM: What makes Computer genuinely useful for planning work: it holds the full context and adjusts everything when one thing changes.

Share Your Work

Once you're happy with the landing page, hit Share to send it directly to teammates by email, copy the link, or enable Allow remix so others can create their own version without touching yours.

[7/2, 10:58 PM] OUR DAILY WISDOM: Beyond Trip Planning

The same logic works beyond trip planning:

Files → risk list: "Review this survey export and flag the top three planning risks"

Content pipeline: "Turn the offsite research into a draft internal announcement"

Recurring reports: "Every Monday, summarize budget changes, open decisions, and next steps"

[7/2, 11:01 PM] OUR DAILY WISDOM: Beyond Trip Planning

The same logic works beyond trip planning:

Files → risk list: "Review this survey export and flag the top three planning risks"

Content pipeline: "Turn the offsite research into a draft internal announcement"

Recurring reports: "Every Monday, summarize budget changes, open decisions, and next steps"

[7/2, 11:04 PM] OUR DAILY WISDOM: Beyond Trip Planning

The same logic works beyond trip planning:

Files → risk list: "Review this survey export and flag the top three planning risks"

Content pipeline: "Turn the offsite research into a draft internal announcement"

Recurring reports: "Every Monday, summarize budget changes, open decisions, and next steps"

From Research to Deliverable

Put the Computer workflow in the right order.

From Research to Deliverable

Put the Computer workflow in the right order.

Drag the items below into order

Draggable item f was dropped over droppable area d

Incorrect

Almost there! Review the steps and try again

[7/2, 11:05 PM] OUR DAILY WISDOM: Beyond Trip Planning

The same logic works beyond trip planning:

Files → risk list: "Review this survey export and flag the top three planning risks"

Content pipeline: "Turn the offsite research into a draft internal announcement"

Recurring reports: "Every Monday, summarize budget changes, open decisions, and next steps"

From Research to Deliverable

Put the Computer workflow in the right order.

From Research to Deliverable

Put the Computer workflow in the right order.

Drag the items below into order

Draggable item f was dropped over droppable area d

See answer

Correct order: b, a, d, c, e, f

[7/2, 11:06 PM] OUR DAILY WISDOM: Key Takeaways

Computer is a separate workspace: It's not a search mode or a research feature.

Use it when the task has multiple steps and a specific deliverable: a plan, a brief, a document.

Workflows give you ready-made task templates: Find them under Customise.

Review before you share: Flag assumptions, verify key figures, separate confirmed from open.

Computer holds full context: Update one constraint and it revises the whole deliverable.

[7/5, 11:34 PM] OUR DAILY WISDOM: Skills for Repeatable Workflows

Every repeated task comes with a hidden cost: rewriting the same instructions, reformatting the same output, running the same checks. Skills eliminate that. Store the process once, and Computer follows it every time you need that kind of work done.

In this lesson, you'll learn how to use Skills to your advantage. Let's go!

[7/5, 11:35 PM] OUR DAILY WISDOM: What Are Skills?

Skills are reusable capabilities for Perplexity Computer — step-by-step instructions it can automatically apply whenever a matching task comes up.

You may have seen something similar in Claude, where Skills standardize how it writes or responds. Computer Skills go further: they're full automated workflows that Computer can re-run across models and tools, turning your instructions into a production-ready capability.

A Skill can store anything Computer needs to approach a type of task consistently:

Structure and workflow steps

Tone and formatting rules

Review criteria and quality checks

Brand standards and assets

Output requirements

[7/5, 11:36 PM] OUR DAILY WISDOM: Once a Skill is set up, Computer applies it every time that kind of work comes up. You won't need to re-explain the task, context, or requirements.

[7/5, 11:37 PM] OUR DAILY WISDOM: What do you think a Skill stores that a one-time prompt doesn't?

A set of source preferences so Computer always pulls from the same places

The repeatable process behind a type of work — structure, steps, and review criteria

The topic and context of the task, so Computer doesn't need background information each time

Incorrect answer

Context and source preferences are useful in a prompt, but they don't make a process repeatable. A Skill defines how Computer approaches a type of work — not just what information it starts with.

[7/5, 11:38 PM] OUR DAILY WISDOM: Where to Find Skills

Open Customise in the sidebar and click the Skills tab. You'll see three views: All, My skills, and Example skills. Example Skills are pre-built templates you can use as-is or as a starting point for your own.

[7/5, 11:46 PM] OUR DAILY WISDOM: Each Skill has a name, a description of when to use it, and a SKILL.md file with the full instructions. Open any example to see what a complete Skill looks like before you build your own.

[7/5, 11:46 PM] OUR DAILY WISDOM: Creating a Skill

Click + Create skill in the top right. You'll see two options: "Create with Computer" builds the Skill through a conversation, and "Upload a skill" lets you import an existing .md or .zip file.

[7/5, 11:47 PM] OUR DAILY WISDOM: Create with Computer" is the easiest starting point. Click the option, and Perplexity already has a head start — it uses its own built-in "create-skill" Skill to open the conversation. Computer asks you questions about the task, the context and intended user, and what a good result looks like. You answer, and it builds the Skill from there.

[7/5, 11:48 PM] OUR DAILY WISDOM: Upload a skill is useful when instructions already exist somewhere else: a process doc, a team playbook, or a Skill from another platform. Export it as a .md file or a .zip and import it directly.

[7/5, 11:49 PM] OUR DAILY WISDOM: Your colleague used the same Skill to produce three competitor analysis reports. Each one came back with the same structure and review checks. What made that possible?

The reports were similar enough that Computer defaulted to the same output format

The Skill defined the structure and review process, so Computer followed the same method each time

Computer remembered the format from the first report and applied it automatically

Correct answer

That's it! Computer doesn't carry memory between tasks — the Skill is what kept the output consistent. It defined the method once, and Computer applied it every time.

[7/5, 11:49 PM] OUR DAILY WISDOM: Query vs. Skill

To use Skills well, it helps to know what they do that a query doesn't. A query gives Computer the assignment. A Skill gives it the method.

Query: "Create a planning brief for the Budapest offsite" — the task right now.

Skill: "Whenever you create an offsite planning brief, include a recommendation, schedule, budget risks, flagged assumptions, open questions, and next steps" — the repeatable process behind every brief.

The query changes with each new destination. The Skill keeps the method stable. Budapest, Riga, Lisbon — same structure, same review checks, every time.

[7/5, 11:51 PM] OUR DAILY WISDOM: Query or Skill?

Decide whether this is a query or a Skill.

"Whenever you write a research brief, use this structure: executive summary, key findings, source list, and open questions."

Query

Skill

Amazing!

Exactly! This defines how Computer should approach not just one task, but the method behind every research brief. That's a Skill.

[7/5, 11:52 PM] OUR DAILY WISDOM: Skill or One-Time Task?

Decide whether this task is worth turning into a Skill.

Writing a competitor analysis report with the same sections and evaluation criteria each month

Worth a Skill

One-time task

Amazing!

Right on! The task repeats, and the process should stay consistent. That's exactly what a Skill is built for: store the method once, apply it every time.

[7/5, 11:53 PM] OUR DAILY WISDOM: Building Your First Skill

Now let's put it into practice. The offsite planning process you've built across this course is a strong Skill candidate. It needs the same structure, sections, and review checks every time a new destination comes up.

You can either answer the questions Perplexity asks when you click "Create with Computer" or come up with the prompt yourself.

[7/5, 11:53 PM] OUR DAILY WISDOM: Create the Skill

Ask Computer to turn your offsite planning process into a reusable Skill.

Perplexity

Create

an offsite planning Skill

that

uses

[inputs]

follows

comparing logistics and flagging risks

and

produces

[output format]

Almost right

Almost there! Review the steps and try again

[7/5, 11:54 PM] OUR DAILY WISDOM: Create the Skill

Ask Computer to turn your offsite planning process into a reusable Skill.

Perplexity

Create an offsite planning Skill that uses destination research, budget notes, and team constraints, follows comparing logistics and flagging risks, and produces a recommendation, schedule, budget risks, open questions, and next steps

I'll structure this as a reusable Skill file rather than a one-off planning template, so it can guide future offsite planning work consistently. First, I'm loading the Skill-creation guidance, then I'll outline the build plan for your approval.

[7/5, 11:54 PM] OUR DAILY WISDOM: Computer doesn't just take your answers and save them — it builds the Skill properly. It loads its own Skill-creation guidance, outlines a build plan for you to approve, then works through several steps before saving the finished Skill. You stay in control at every stage.

[7/5, 11:55 PM] OUR DAILY WISDOM: Great! Once your Skill is saved, you’ll find it under "My skills."

[7/5, 11:55 PM] OUR DAILY WISDOM: Skills With Brand Guidelines

One of the most practical uses of Skills is brand consistency. A Skill can store your organization's brand guidelines — tone, colors, fonts, formatting rules — so Computer applies them to any deliverable it produces.

When you upload a Skill as a .zip file, you can include supporting assets alongside the instructions: font files, color palettes, logo variations, and other brand files. Computer uses them when building presentations, documents, or any other branded output.

[7/5, 11:56 PM] OUR DAILY WISDOM: Remember the Prague offsite landing page Computer built in the previous lesson? Instead of adjusting it manually, you apply a brand Skill, and Computer realigns every slide to your organization's standards in one task.

[7/5, 11:57 PM] OUR DAILY WISDOM: applies them to any deliverable it produces.

Remember the Prague offsite landing page Computer built in the previous lesson? Instead of adjusting it manually, you apply a brand Skill, and Computer realigns every slide to your organization's standards in one task.

Apply the Brand Skill

Ask Computer to update a landing page using your brand guidelines Skill.

Apply the Brand Skill

Ask Computer to update a landing page using your brand guidelines Skill.

Perplexity

Use the brand-guidelines Skill to update the Prague offsite landing page. Align fonts, colors, and slide layouts with brand standards and flag anything that needs manual review.

Updated and redeployed the Prague offsite landing page using the brand-guideline skill.

What changed:

— Fonts now follow brand standards: Satoshi for headlines/display, Inter for body/UI.

— Colors now follow brand standards: white/light-gray base, Electric Blue CTAs, Neon Lime used only as small status/section accents.

— Removed the previous Prague terracotta/serif visual system.

Manual review flags:

— "Slide layouts" do not directly apply to this web landing page unless there is a related deck to update.

— The custom Prague rooftop logo is still event-specific. Review if the company requires the official brand mark instead.

— RSVP copy and agenda content are unchanged beyond visual brand alignment. Review if you want stricter brand voice editing.

Amazing!

You're right on track with your approach

[7/5, 11:57 PM] OUR DAILY WISDOM: This is how the landing page changed with the brand-guideline skill!

When the Output Drifts

Skills keep output consistent, but they need updating when results drift. If Computer skips a section or handles something differently, update the Skill, not just the output.

[7/5, 11:59 PM] OUR DAILY WISDOM: Computer produces a competitor report using your Skill, but it skips the recommendations section and doesn't flag data older than 12 months. What should you do?

Select all possible answers

Fix just this report manually and move on

Update the Skill to require a recommendations section every time

Add a review check to the Skill for data recency

Ask Computer to revise the report using the updated Skill

Incorrect answer

A one-off fix disappears after this report. The Skill is where the process lives — updating it once is what prevents the same drift from happening next time.

Skills for Teams

A Skill you build doesn't have to stay personal. Once it's saved, you can share it with your team so everyone follows the same process, not just you.

This is where Skills become a team standard rather than a personal shortcut.

Beyond One Task

The same logic applies across any repeated work where consistency matters:

Weekly research summaries: Same structure, sources, and format each time

Competitor comparison reports: Same sections and evaluation criteria

Brand-compliant assets: Fonts, colors, and layouts applied automatically

CSV or budget analysis: Same review checks and risk flags on every file

[7/5, 11:59 PM] OUR DAILY WISDOM: Skills for Repeatable Workflows

In this lesson, you'll learn how to use Skills to your advantage. Let's go!

What Are Skills?

Skills are reusable capabilities for Perplexity Computer — step-by-step instructions it can automatically apply whenever a matching task comes up.

A Skill can store anything Computer needs to approach a type of task consistently:

Structure and workflow steps

Tone and formatting rules

Review criteria and quality checks

Brand standards and assets

Output requirements

Once a Skill is set up, Computer applies it every time that kind of work comes up. You won't need to re-explain the task, context, or requirements.

What do you think a Skill stores that a one-time prompt doesn't?

A set of source preferences so Computer always pulls from the same places

The repeatable process behind a type of work — structure, steps, and review criteria

The topic and context of the task, so Computer doesn't need background information each time

Incorrect answer

Context and source preferences are useful in a prompt, but they don't make a process repeatable. A Skill defines how Computer approaches a type of work — not just what information it starts with.

Where to Find Skills

Each Skill has a name, a description of when to use it, and a SKILL.md file with the full instructions. Open any example to see what a complete Skill looks like before you build your own.

Creating a Skill

Click + Create skill in the top right. You'll see two options: "Create with Computer" builds the Skill through a conversation, and "Upload a skill" lets you import an existing .md or .zip file.

"Create with Computer" is the easiest starting point. Click the option, and Perplexity already has a head start — it uses its own built-in "create-skill" Skill to open the conversation. Computer asks you questions about the task, the context and intended user, and what a good result looks like. You answer, and it builds the Skill from there.

Upload a skill is useful when instructions already exist somewhere else: a process doc, a team playbook, or a Skill from another platform. Export it as a .md file or a .zip and import it directly.

Your colleague used the same Skill to produce three competitor analysis reports. Each one came back with the same structure and review checks. What made that possible?

The reports were similar enough that Computer defaulted to the same output format

The Skill defined the structure and review process, so Computer followed the same method each time

Computer remembered the format from the first report and applied it automatically

Correct answer

That's it! Computer doesn't carry memory between tasks — the Skill is what kept the output consistent. It defined the method once, and Computer applied it every time.

Query vs. Skill

To use Skills well, it helps to know what they do that a query doesn't. A query gives Computer the assignment. A Skill gives it the method.

Query: "Create a planning brief for the Budapest offsite" — the task right now.

The query changes with each new destination. The Skill keeps the method stable. Budapest, Riga, Lisbon — same structure, same review checks, every time.

completed check

Task completed

Query or Skill?

Decide whether this is a query or a Skill.

completed check

Task completed

Skill or One-Time Task?

Decide whether this task is worth turning into a Skill.

Building Your First Skill

You can either answer the questions Perplexity asks when you click "Create with Computer" or come up with the prompt yourself.

completed check

Task completed

Create the Skill

Ask Computer to turn your offsite planning process into a reusable Skill.

Computer doesn't just take your answers and save them — it builds the Skill properly. It loads its own Skill-creation guidance, outlines a build plan for you to approve, then works through several steps before saving the finished Skill. You stay in control at every stage.

Great! Once your Skill is saved, you’ll find it under "My skills."

Skills With Brand Guidelines

completed check

Task completed

Apply the Brand Skill

Ask Computer to update a landing page using your brand guidelines Skill.

This is how the landing page changed with the brand-guideline skill!

When the Output Drifts

Skills keep output consistent, but they need updating when results drift. If Computer skips a section or handles something differently, update the Skill, not just the output.

Computer produces a competitor report using your Skill, but it skips the recommendations section and doesn't flag data older than 12 months. What should you do?

Select all possible answers

Fix just this report manually and move on

Update the Skill to require a recommendations section every time

Add a review check to the Skill for data recency

Ask Computer to revise the report using the updated Skill

Incorrect answer

A one-off fix disappears after this report. The Skill is where the process lives — updating it once is what prevents the same drift from happening next time.

Skills for Teams

A Skill you build doesn't have to stay personal. Once it's saved, you can share it with your team so everyone follows the same process, not just you.

This is where Skills become a team standard rather than a personal shortcut.

Beyond One Task

The same logic applies across any repeated work where consistency matters:

Weekly research summaries: Same structure, sources, and format each time

Competitor comparison reports: Same sections and evaluation criteria

Brand-compliant assets: Fonts, colors, and layouts applied automatically

CSV or budget analysis: Same review checks and risk flags on every file

Key Takeaways

Skills live in Customise → Skills: Open any Example Skill to see what a complete one looks like.

Query tells Computer what to do; Skill tells it how to approach that kind of work: The distinction isn't just a task, but a method.

Create with Computer: Describe the task and let Computer build the Skill through conversation.

Upload as .zip to include brand assets: Fonts, color palettes, and files Computer can use in output.

Computer can recognize when a Skill fits: It may apply a relevant Skill automatically.

Skills scale across teams: One Skill built means one consistent process shared.

[7/6, 12:00 AM] OUR DAILY WISDOM: What's Next?

Skills are your last customization layer inside Computer. Next, you'll learn about Connectors — how to link Perplexity to the tools and data sources you already use, so it works with your actual workflow, not just the open web.

[7/11, 3:31 PM] OUR DAILY WISDOM: Your AI Still Doesn't Know You

You've built a Space, handed workflows to Computer, and created Skills for repeated tasks. But every time a new task starts, you're still the one copying in the latest email, the updated budget, and the calendar conflict. Computer is capable, but it doesn't have access to where the work actually lives. That's what Connectors change.

In this lesson, you'll connect Computer to the tools you already use, learn when it can take action on your behalf, and set up a workflow that runs automatically.

[7/11, 3:32 PM] OUR DAILY WISDOM: Your AI Still Doesn't Know You

In this lesson, you'll connect Computer to the tools you already use, learn when it can take action on your behalf, and set up a workflow that runs automatically.

What Are Connectors?

Connectors link Perplexity Computer to the apps where your work lives. Without them, Computer only knows what you paste in. With them, it can pull from your inbox, calendar, files, and project tools, and work with the current state of your work, not a snapshot you manually assembled.

[7/11, 3:34 PM] OUR DAILY WISDOM: Your team lead emails a new constraint for the Prague offsite. Your coordinator updates a checklist in Notion. Two people mark themselves unavailable on Friday in Google Calendar. Without Connectors, you copy all of that in yourself, but with them, Computer reads it directly and works from there!

Where to Find Connectors

Open Customise in the sidebar and click the Connectors tab. You'll see four views: Discover, All, Connected, and Available. Some connectors are included with Computer at no extra cost — others are available to add

[7/11, 3:36 PM] OUR DAILY WISDOM: Click any connector to see what it can do: capabilities listed on the left, tools broken into read-only and write categories on the right. Click Add connector to link your account.

[7/11, 3:37 PM] OUR DAILY WISDOM: Once connected, the app appears under Connected with your account shown. It stays linked until you disconnect it manually.

[7/11, 3:39 PM] OUR DAILY WISDOM: You connect Gmail to Computer and ask it to find the latest email from your team lead. What has actually changed about how Computer works?

Computer can now send emails on your behalf without additional instructions

Computer can now search the web using your email history as additional context

Computer will automatically monitor your inbox and flag important emails going forward

Computer no longer needs you to paste in context — it can read from where the information lives

Incorrect answer

Connecting an app doesn't change how Computer searches the web, trigger automatic monitoring, or grant send permissions. It changes where Computer can look for context from what you paste in, to where the information lives.

[7/11, 3:42 PM] OUR DAILY WISDOM: Read vs. Action

Every connector has two tool types: read-only and write.

Read tools let Computer pull context: searching emails, checking availability, reviewing files.

Write tools let Computer prepare actions: drafting a reply, updating a task, creating a calendar event.

The key difference: read actions happen silently. Write actions always need your approval before they run. Computer shows you the draft or proposed change first — you decide whether it goes ahead.

[7/11, 3:45 PM] OUR DAILY WISDOM: Needs Connectors or Not?

Decide whether this task is a good use of Connectors.

Asking Computer to summarize a document you just uploaded to the chat

Needs Connectors

Doesn't need Connectors

Incorrect

Connectors connect Computer to external tools. If the information is already in the conversation, there's nothing to connect to — the task works fine without them.

[7/11, 3:46 PM] OUR DAILY WISDOM: Needs Connectors or Not?

Decide whether this task is a good use of Connectors.

Asking Computer to summarize a document you just uploaded to the chat

Needs Connectors

Doesn't need Connectors

Amazing!

Right! The document is already in the chat — Computer has what it needs. Connectors earn their place when the context lives somewhere else: in an inbox, a calendar, a project tracker, or a CRM.

[7/11, 3:47 PM] OUR DAILY WISDOM: The key difference: read actions happen silently. Write actions always need your approval before they run. Computer shows you the draft or proposed change first — you decide whether it goes ahead.

completed check

Task completed

Needs Connectors or Not?

Decide whether this task is a good use of Connectors.

Pull Context From Your Tools

Ask Computer to gather the latest offsite planning context from connected apps.

Pull Context From Your Tools

Ask Computer to gather the latest offsite planning context from connected apps.

Perplexity

Check Gmail for the latest offsite update from the team lead, review Google Calendar for availability conflicts, and create a short planning summary with open decisions flagged

Offsite Planning Summary

Team lead confirmed offsite cannot start before Wednesday afternoon.

Calendar conflicts

Two team members unavailable Friday morning — Day 3 morning session needs rescheduling.

Open decisions: hotel deposit deadline, dietary requirements, and final arrival times are still unconfirmed.

Amazing!

You're right on track with your approach

[7/11, 3:50 PM] OUR DAILY WISDOM: The key difference: read actions happen silently. Write actions always need your approval before they run. Computer shows you the draft or proposed change first — you decide whether it goes ahead.

completed check

Task completed

Needs Connectors or Not?

Decide whether this task is a good use of Connectors.

Pull Context From Your Tools

Ask Computer to gather the latest offsite planning context from connected apps.

Pull Context From Your Tools

Ask Computer to gather the latest offsite planning context from connected apps.

Perplexity

Check Gmail for the latest offsite update from the team lead, review Google Calendar for availability conflicts, and create a short planning summary with open decisions flagged

Offsite Planning Summary

Team lead confirmed offsite cannot start before Wednesday afternoon.

Calendar conflicts

Two team members unavailable Friday morning — Day 3 morning session needs rescheduling.

Open decisions: hotel deposit deadline, dietary requirements, and final arrival times are still unconfirmed.

Amazing!

You're right on track with your approach

[7/11, 3:51 PM] OUR DAILY WISDOM: Computer read both tools, combined the context, and surfaced what still needs a decision. All that from one prompt!

When Computer Takes Action

Some connectors go beyond reading. Computer can draft emails, update tasks, or create calendar events — but it shows you what it intends to do before anything changes.

Let's see how that works in the next task.

[7/11, 3:53 PM] OUR DAILY WISDOM: Draft Before Sending

Ask Computer to find a newsletter in your inbox and prepare an unsubscribe request for your review.

Perplexity

Find the most recent newsletter email in my Gmail, summarize it in 2 sentences, and draft an unsubscribe request back to the sender

Done. I found the latest newsletter-style email: Weekly Growth Digest, received Tuesday.

Summary: roundup of SaaS growth tactics and tool recommendations.

I created a reply draft asking them to unsubscribe you from newsletters and promotional emails. Want me to find the direct unsubscribe link in the email too?

Amazing!

You're right on track with your approach

[7/11, 3:54 PM] OUR DAILY WISDOM: Computer found the email, wrote the reply, and even created a draft right in your Gmail. The only thing left for you to do is to review it and hit the send button. That's how writing tools work: Computer prepares, and you decide.

[7/11, 3:56 PM] OUR DAILY WISDOM: Your AI Still Doesn't Know You

In this lesson, you'll connect Computer to the tools you already use, learn when it can take action on your behalf, and set up a workflow that runs automatically.

What Are Connectors?

Your team lead emails a new constraint for the Prague offsite. Your coordinator updates a checklist in Notion. Two people mark themselves unavailable on Friday in Google Calendar. Without Connectors, you copy all of that in yourself, but with them, Computer reads it directly and works from there!

Where to Find Connectors

Click any connector to see what it can do: capabilities listed on the left, tools broken into read-only and write categories on the right. Click Add connector to link your account.

Once connected, the app appears under Connected with your account shown. It stays linked until you disconnect it manually.

You connect Gmail to Computer and ask it to find the latest email from your team lead. What has actually changed about how Computer works?

Computer can now send emails on your behalf without additional instructions

Computer no longer needs you to paste in context — it can read from where the information lives

Computer can now search the web using your email history as additional context

Computer will automatically monitor your inbox and flag important emails going forward

Incorrect answer

Read vs. Action

Every connector has two tool types: read-only and write.

Read tools let Computer pull context: searching emails, checking availability, reviewing files.

Write tools let Computer prepare actions: drafting a reply, updating a task, creating a calendar event.

The key difference: read actions happen silently. Write actions always need your approval before they run. Computer shows you the draft or proposed change first — you decide whether it goes ahead.

completed check

Task completed

Needs Connectors or Not?

Decide whether this task is a good use of Connectors.

completed check

Task completed

Pull Context From Your Tools

Ask Computer to gather the latest offsite planning context from connected apps.

Computer read both tools, combined the context, and surfaced what still needs a decision. All that from one prompt!

When Computer Takes Action

Some connectors go beyond reading. Computer can draft emails, update tasks, or create calendar events — but it shows you what it intends to do before anything changes.

Let's see how that works in the next task.

completed check

Task completed

Draft Before Sending

Ask Computer to find a newsletter in your inbox and prepare an unsubscribe request for your review.

Computer found the email, wrote the reply, and even created a draft right in your Gmail. The only thing left for you to do is to review it and hit the send button. That's how writing tools work: Computer prepares, and you decide.

Computer drafts an email to your coordinator with the revised offsite schedule. What should you do before it sends?

Let it send — the plan is finalized, and the coordinator is waiting

Review the draft and approve before it goes out

Copy the draft manually and send it yourself to be safe

Disconnect Gmail so Computer can't access email

Correct answer

Nailed it! Connected doesn't mean uncontrolled. Review every draft, update, and action before it leaves Computer — that's the habit that keeps connected workflows safe.

[7/11, 3:58 PM] OUR DAILY WISDOM: Connectors in Everyday Work

The offsite scenario is one example. The same logic applies anywhere work is spread across tools.

[7/11, 4:00 PM] OUR DAILY WISDOM: Your AI Still Doesn't Know You

In this lesson, you'll connect Computer to the tools you already use, learn when it can take action on your behalf, and set up a workflow that runs automatically.

What Are Connectors?

Where to Find Connectors

Click any connector to see what it can do: capabilities listed on the left, tools broken into read-only and write categories on the right. Click Add connector to link your account.

Once connected, the app appears under Connected with your account shown. It stays linked until you disconnect it manually.

You connect Gmail to Computer and ask it to find the latest email from your team lead. What has actually changed about how Computer works?

Computer will automatically monitor your inbox and flag important emails going forward

Computer can now send emails on your behalf without additional instructions

Computer no longer needs you to paste in context — it can read from where the information lives

Computer can now search the web using your email history as additional context

Correct answer

That's it! The shift isn't just one email — it's that Computer now works from your actual tools instead of whatever you remembered to include. That's what makes connected workflows fundamentally different.

Read vs. Action

Every connector has two tool types: read-only and write.

Read tools let Computer pull context: searching emails, checking availability, reviewing files.

Write tools let Computer prepare actions: drafting a reply, updating a task, creating a calendar event.

The key difference: read actions happen silently. Write actions always need your approval before they run. Computer shows you the draft or proposed change first — you decide whether it goes ahead.

completed check

Task completed

Needs Connectors or Not?

Decide whether this task is a good use of Connectors.

completed check

Task completed

Pull Context From Your Tools

Ask Computer to gather the latest offsite planning context from connected apps.

Computer read both tools, combined the context, and surfaced what still needs a decision. All that from one prompt!

When Computer Takes Action

Some connectors go beyond reading. Computer can draft emails, update tasks, or create calendar events — but it shows you what it intends to do before anything changes.

Let's see how that works in the next task.

completed check

Task completed

Draft Before Sending

Ask Computer to find a newsletter in your inbox and prepare an unsubscribe request for your review.

Computer drafts an email to your coordinator with the revised offsite schedule. What should you do before it sends?

Copy the draft manually and send it yourself to be safe

Disconnect Gmail so Computer can't access email

Let it send — the plan is finalized, and the coordinator is waiting

Review the draft and approve before it goes out

Correct answer

Nailed it! Connected doesn't mean uncontrolled. Review every draft, update, and action before it leaves Computer — that's the habit that keeps connected workflows safe.

[7/11, 4:01 PM] OUR DAILY WISDOM: Connectors in Everyday Work

The offsite scenario is one example. The same logic applies anywhere work is spread across tools.

Inbox management: "Find all newsletters I haven't opened in 30 days and draft unsubscribe requests for each."

Meeting prep: "Check my calendar for tomorrow, find related emails for each meeting, and create a one-line brief for each one."

Project tracking: "Pull open tasks from Asana, check if any are mentioned in recent Slack threads, and flag what's overdue."

CRM updates: "Review this week's client emails in Gmail and draft follow-up notes for each contact in HubSpot."

[7/11, 4:02 PM] OUR DAILY WISDOM: Scheduled Tasks

Every workflow above runs on demand: Computer works when you ask it. Scheduled tasks let Computer run a connected workflow automatically, on a cadence you set.

[7/11, 4:02 PM] OUR DAILY WISDOM: Scheduled Tasks

Every workflow above runs on demand: Computer works when you ask it. Scheduled tasks let Computer run a connected workflow automatically, on a cadence you set.

For the offsite: every Monday morning, check the planning Space, coordinator notes, and latest emails — then produce a short update with what changed, what's blocked, and what needs a decision this week. Set it once, and it runs without you.

[7/11, 4:04 PM] OUR DAILY WISDOM: Schedule the Update

Set up a recurring offsite planning summary using connected tools.

Perplexity

Every

Monday morning

check

[sources]

and

send

[output]

with changes, blockers, and decisions needed

Almost right

Almost there! Review the steps and try again

[7/11, 4:05 PM] OUR DAILY WISDOM: Schedule the Update

Set up a recurring offsite planning summary using connected tools.

Perplexity

Every Monday morning, check the Team Offsite Space, Gmail, and the coordinator checklist, and send a short update with changes, blockers, and decisions needed

[7/11, 4:06 PM] OUR DAILY WISDOM: Before scheduling, Computer asks you a few setup questions — what time to run, where the sources are hosted, and how to deliver the update. Answer them once, and it handles the rest automatically.

And because scheduled tasks run in the cloud, they don't need your device to be on.

Connect Only What You Need

Remember: more access isn't always better. Before connecting an app, ask whether Computer actually needs it for the task. Good connected workflows use the minimum context needed — not everything that's available.

[7/11, 4:07 PM] OUR DAILY WISDOM: Before any connected workflow, run through these:

Which apps hold the context this task really needs?

Should Computer read only, or prepare actions too?

What should always require my approval before it changes?

Does this task repeat enough to be worth scheduling?

[7/11, 4:10 PM] OUR DAILY WISDOM: Build a Connected Workflow

Put the connector workflow steps in the right order.

Drag the items below into order

Draggable item d was dropped over droppable area b

See answer

Correct order: c, b, e, a, d

[7/11, 4:11 PM] OUR DAILY WISDOM: Build a Connected Workflow

Put the connector workflow steps in the right order.

Drag the items below into order

Draggable item d was dropped over droppable area b

See answer

Correct order: c, b, e, a, d

[7/11, 4:11 PM] OUR DAILY WISDOM: Key Takeaways

Connectors live in "Customise": Browse by Discover, All, Connected, or Available.

Without Connectors, Computer works from what you paste in: With them, it works from where the information lives.

Read tools pull context silently; write tools prepare actions for your approval.

Tell Computer where to look: Name the tools, what to find, and what output to produce.

Scheduled tasks run connected workflows automatically: Set the cadence once, and Computer handles the check-in

[7/11, 4:12 PM] OUR DAILY WISDOM: What's Next?

You've now connected the tools around your workflow. Next, you'll put the full Perplexity toolkit to work across real use cases — research, reporting, and more!

[7/12, 7:58 PM] OUR DAILY WISDOM: News & Trend Analysis

You wake up to a flood of messages about something that happened overnight — a platform changed its algorithm, a brand said something it shouldn't have, a market shifted. Everyone is reacting. You want to understand what happened.

In this lesson, you'll learn how to track breaking news, compare sources, and turn a research session into a structured report you can act on.

[7/12, 8:00 PM] OUR DAILY WISDOM: You open the web to follow a breaking story. There are dozens of results from different outlets and different time periods. What's the most important thing to pay attention to?

How many sources cover the same angle — consensus signals the information is reliable

When each result was published, since stories change fast

Which outlet published it — bigger publications are more accurate on breaking news

Incorrect answer

Outlet reputation and source consensus both matter for credibility — but neither tells you where a result sits in the news cycle. A well-known outlet's coverage from three days ago may already be outdated, and consensus across sources can simply mean the same early report got widely picked up. Date is what orients you first.

[7/12, 8:01 PM] OUR DAILY WISDOM: Searching for a breaking story without filters is like opening a firehose. You get opinion pieces, outdated takes, and reactions to reactions — mixed in with the actual facts. The story gets harder to understand.

Perplexity makes that easier: you can filter by date, compare perspectives, and verify claims in one place instead of switching between tabs. The first thing to add to any breaking news search is a date filter.

[7/12, 8:02 PM] OUR DAILY WISDOM: Tracking Breaking News

You learned how date operators work in Lesson 5. Here, they do the same job in a news context — restricting results to a specific window, whether that's the last 24 hours, last week, or last month. In a fast-moving story, that single addition keeps the results focused on today's development. Let's apply it to a news search.

[7/12, 8:03 PM] OUR DAILY WISDOM: — Prioritize original, platform‑native Reels and strong hooks.

— Use fewer, more specific hashtags (3-5) to signal clear topics.

— Focus on watch time, comments, shares, and profile visits rather than just posting volume.

How to find the official announcement you're tracking

If you're specifically looking for Instagram's own public announcement page dated after 2026‑03‑03, the most reliable sources are:

— The Instagram Creators section of the Meta Newsroom and Instagram's official blog.

— Meta's "Creator Updates" newsletter and Instagram's @creators account, which highlight new algorithmic tweaks and feature rollouts.

Sources: TheBusinesNews, InfinityDigitalSpace

Amazing!

You're right on track with your approach

[7/12, 8:04 PM] OUR DAILY WISDOM: The search result pointed you to Instagram's official newsroom to confirm the announcement. What should you do?

Follow it, that's where the original announcement lives

Ignore it, the search already returned current coverage

Note it and come back later, the search result is enough to start with

Correct answer

The newsroom suggestion is Perplexity pointing you to where the claim originated. What follows makes the result citable, not just useful.

[7/12, 8:05 PM] OUR DAILY WISDOM: Perplexity pointing you to additional sources is the starting point for verification. One search gives you the shape of the story. Cross-checking the primary source is what confirms it. That's what the next step covers.

[7/12, 8:06 PM] OUR DAILY WISDOM: Comparing Perspectives

The same story reads differently depending on who's telling it. Instagram's newsroom says the change helps creators. Analysts say it disrupts brands. Creators report mixed results on the ground. None of them is wrong; they're just seeing different parts of the same thing. So let's try pulling from more than one source.

[7/12, 8:07 PM] OUR DAILY WISDOM: Multi-Source Perspective Search

Search for reactions to the same algorithm change across different source types — platform, creators, and industry analysts.

Perplexity

Instagram algorithm change after:2026-03-03, perspectives from Instagram's own newsroom and industry analysts

Instagram's own perspective

Instagram frames the 2026 algorithm as more modular and user‑controlled, not a single "ranking formula" for everything.

"Your Algorithm" for Reels (launched late 2025, expanding in 2026):

Users can now tell Instagram which topics they want to see more or less of in Reels, and the system uses that as a primary ranking signal.

Meta’s official line is that this gives people more control and pushes creators to match explicitly chosen interests, not just assumed preferences.

Industry analysts see three main shifts:

More search‑ and topic‑driven discovery (content must be "searchable" and hook cold audiences).

Stronger bias toward well‑retained Reels and vertical carousels over static posts.

Lower reach for "for‑followers‑only" riving hooks.

Amazing!

You're right on track with your approach

[7/12, 8:09 PM] OUR DAILY WISDOM: The divergence between Instagram's framing and the analyst's view is where the story lies. Platform announcements protect the platform. Analyst coverage reflects market implications.

For any story with multiple affected groups, the gap between perspectives is what you need to understand before acting.

Discover — News Without Searching

Aside from searching directly, Perplexity has a "Discover" section at the top of the screen that surfaces trending news without any prompt.

It includes For You, Top, and topic categories like Tech & Science, Entertainment, Sports, and Business — useful for staying on top of stories before they hit your inbox.

[7/12, 8:10 PM] OUR DAILY WISDOM: Checking What You Find

Breaking news move fast — and so do unverified claims. A citation tells you where a claim came from, not whether it held up. One result claims that accounts under 10K are seeing a 40% increase in reach. Before using that figure, it needs a check.

[7/12, 8:11 PM] OUR DAILY WISDOM: screen that surfaces trending news without any prompt.

It includes For You, Top, and topic categories like Tech & Science, Entertainment, Sports, and Business — useful for staying on top of stories before they hit your inbox.

Checking What You Find

Verifying Data

Verify the information before including it in anything you publish or present.

Verifying Data

Verify the information before including it in anything you publish or present.

Perplexity

Instagram algorithm update after:2026-03-03. Verify the claim: accounts under 10K see 40% reach increase. Check primary sources and platform data.

There is no verifiable primary‑source support for the specific claim that "accounts under 10K followers see a 40% reach increase after the March 2026 Instagram algorithm update." Neither Instagram's official Newsroom / Instagram Blog nor major platform‑analytics providers (e.g., Hootsuite, Sprout Social, Buffer) publish that exact statistic.

How that claim is likely circulating

The "sub‑10K +40% reach" line appears to be anecdotal or speculative rather than a published metric from Instagram or a reputable analytics vendor:

— Third‑party blogs and videos often generalize when discussing "small creators benefitting" from newer interest‑based, recommendation‑driven ranking, but they do not cite Instagram‑provided reach‑lift data by follower bracket.

— Some industry reports talk about better discoverability for smaller accounts in Explore and Reels due to topic‑matching and retention‑based signals, but they avoid precise percentage lifts such as "40%."

Sources: Techwyse, Inro, TheBusinessNews

Amazing!

You're right on track with your approach

[7/12, 8:12 PM] OUR DAILY WISDOM: See how Perplexity traced the claim back to third-party blogs generalizing about small creator growth, with no Instagram-provided data behind it. The underlying trend has support: smaller accounts are seeing better discoverability through topic-matching and retention signals.

That's the distinction: you carry the trend into your report, hold off on the figure until primary data backs it.

[7/12, 8:13 PM] OUR DAILY WISDOM: https://app.coursiv.com/guides/3616efef-4575-80a4-a433-f6ed5ec2bb43/f1dcdbc7-172e-4208-8e88-c40f6b97b404/3616efef-4575-81a2-bf73-c228cc37c019?isLastLesson=false&isAudio=false&hasAudio=true

[7/12, 8:13 PM] OUR DAILY WISDOM: Building a Trend Report

You now have current results, multiple perspectives, and a verified account of what actually happened. You can build a structured summary that someone can read and act on without doing the research themselves. A LinkedIn post, a team briefing, or a client update — whatever the format, the research is already done.

[7/12, 8:14 PM] OUR DAILY WISDOM: Building a Trend Report

Structured Trend Report

Compile everything into a single structured report covering what happened, the timeline, perspectives, and implications.

Structured Trend Report

Compile everything into a single structured report covering what happened, the timeline, perspectives, and implications.

Perplexity

Compile a trend report on "Instagram algorithm change": include what happened, timeline, and perspectives from key groups, after:2026-03-03

Instagram algorithm change trend report (post‑March 2026 snapshot)

Instagram's 2026‑style algorithm changes focus on user‑controlled topics, interest‑based discovery, and retention‑driven ranking, not a single labeled "update" after March 2026. Key shifts include stronger ranking by watch time, sends, saves, and profile visits; more emphasis on Reels and Explore; and less reliance on raw likes or follower count.

Platform‑sized tools like Hootsuite, Sprout Social, and Buffer highlight that high‑retention, topic‑clear content now gains more visibility, while brands and creators are shifting toward vertical Reels, topic‑series, and hooks that work for non‑followers. There is no official support for viral claims like "40% reach boost for sub‑10K accounts," only that well‑retained, interest‑matched content benefits most.

Sources: TheBuisnessNews, Buffer, SproutSocial

Amazing!

You're right on track with your approach

[7/12, 8:16 PM] OUR DAILY WISDOM: What makes Perplexity useful for news research?

Select all possible answers

The most credible sources always appear at the top of the results

Claims that lack a primary source get flagged, so you know what to verify

A better-structured prompt makes the underlying data more reliable

Different source types come back in one search instead of opening separate tabs

Date filters keep results focused on what actually happened recently

Incorrect answer

Source order still needs your judgment — credibility isn't automatic. And prompt structure shapes the output, not the data behind it. A false claim in a well-structured prompt is still false. Verification is what catches it, not the prompt.

[7/12, 8:17 PM] OUR DAILY WISDOM: Key Takeaways

Date filters are the first thing you add to any news search: They keep current results from competing with older ones.

Multiple source types surface the full picture: Each one covers different ground.

Discover trending stories without searching: Useful for staying on top of topics before they become urgent.

Structured reports turn a research session into a deliverable: Specify sections, and the output arrives organized.

[7/12, 8:18 PM] OUR DAILY WISDOM: You can now track a breaking story, filter for what's current, check what you find, and turn the research into something your team can act on. But how do you know if a source can actually be trusted in the first place? That's what the next lesson covers.

You'll learn how to evaluate source credibility and spot when information shouldn't be trusted before it ends up in something you've published. On to the next!