What it does that search doesn't

The researcher desk from your tool map gets its own course now. Perplexity's proposition is simple to state: ask a question, get a synthesized answer with the sources it came from attached. This opening lesson is about what that actually changes — versus a search engine on one side and a plain chat on the other — and the mindset that makes sourced answers worth more than either. (Product specifics current as of September 2026.)

Versus search: it reads for you, first. A search engine hands you ten doors; you do the walking. Here, the walking is done — pages read, synthesized, footnoted — and you start from a draft answer instead of a reading list. The win is time; the cost is that someone else chose which doors to open. Both halves of that sentence matter, and the skeptic's habit of glancing at which sources got used never stops paying.

Versus a plain chat: the receipts are native. A generalist chat answers from trained knowledge — brilliant for drafting, blind for the current and the checkable. Here every claim arrives wearing its citation, which transforms your verification workflow: the border crossing you learned in Foundations is one click instead of a separate search. A sourced answer isn't automatically true — but it's automatically checkable, and that difference is the entire product.

Where it slots in your week. Facts that must be current (prices, versions, schedules, news); "says who?" questions where you'll repeat the answer to someone; comparisons that live across many pages; and the background-read before a decision. What stays elsewhere: drafting and voice work (generalist desk), heavy private documents (careful desk), and anything where the answer is in your files rather than the public web.

The mindset shift: from answers to evidence. The habit this desk trains — the one that compounds — is treating every answer as a claim-plus-evidence pair. Fluent prose stops impressing you; the questions become "sourced by whom, dated when, and does the source actually say that?" That's not cynicism, it's professionalism — and it will quietly upgrade how you read everything, AI-generated or not.

Modes exist here too. As of September 2026, the researcher desk has its own gears — a quick sourced answer for everyday lookups, and deeper research runs that read more widely and take minutes (the deep-work lesson later in this course goes properly into those). The three-gear instinct from Foundations transfers whole: match the depth to the cost of being wrong, escalate when the quick answer feels thin, and don't ceremonial-search what one good source settles. Gear names shift with redesigns; the matching rule is yours.

One mistake to skip: using the researcher desk as your everything-desk. Sourced answering is slower and more ceremonial than chat, and pointing it at "rewrite this email" wastes both tools. The map rule holds: this desk is for questions where evidence matters — which is fewer questions than your curiosity generates, and more important ones.