JevMade

Sign in
← Back to guides

JevMade field notes / Research writeup

What should an AI assistant remember?

Dhravya Shah tests ways to help AI assistants sort and save information. Some choices improve search results, but leaving out text can also mean losing details the assistant needs later.

Original by Dhravya ShahRetrieval
Bookmark: What should an AI assistant remember? Keep this in your collection.
Leave a noteWhat would you try with this? : What should an AI assistant remember?

Only you can see your notes.

Listen to this guide

JevMade’s plain-English explanation

0:00 /

Our summary

An AI assistant needs useful information to answer a question. Dhravya Shah, the founder of Supermemory, tests ways to help it find that information. He uses Jev, an AI tool that chooses from options rather than writing an answer.

The tests cover putting useful search results first, splitting long documents into smaller pieces, and choosing which sentences to keep before saving information. The author also explores whether Jev could help an assistant decide when to look up something it has already saved.

The results are mixed. In the author's tests, Jev improved some search results, but other tools sometimes did better or cost less. Splitting documents this way also cost more. Removing sentences saved space, but could leave out details that matter later.

Key takeaways

  1. Compare each job with a simpler way of doing it.
  2. Removing words can save space without making the remaining information more useful.
  3. The idea of choosing when to search saved information still needs testing.

These are the author's findings.

X article · Original published

Read the original guide Opens the author’s site in a new tab.

Your note

Only you can see this. Clear the text and save to remove it.

Keep what you save

Sign in to keep bookmarks and private notes on every device. We email you a link; there is no password, and the same link creates your account.