←back to thread

AI 2027

(ai-2027.com)
949 points Tenoke | 2 comments | 03 Apr 25 16:13 UTC | HN request time: 0.424s | source
Show context
torginus ◴[03 Apr 25 21:44 UTC] No.43575836[source]▶
Much has been made in its article about autonomous agents ability to do research via browsing the web - the web is 90% garbage by weight (including articles on certain specialist topics).

And it shows. When I used GPT's deep research to research the topic, it generated a shallow and largely incorrect summary of the issue, owning mostly to its inability to find quality material, instead it ended up going for places like Wikipedia, and random infomercial listicles found on Google.

I have a trusty Electronics textbook written in the 80s, I'm sure generating a similarly accurate, correct and deep analysis on circuit design using only Google to help would be 1000x harder than sitting down and working through that book and understanding it.

replies(5): >>43576068 #>>43577044 #>>43577133 #>>43581980 #>>43582007 #
somerandomness ◴[03 Apr 25 22:09 UTC] No.43576068[source]▶
Agreed. However, source curation and agents are two different parts of Deep Research. What if you provided that textbook to a reliable agent?

Plug: We built https://RadPod.ai to allow you to do that, i.e. Deep Research on your data.

replies(4): >>43576163 #>>43576409 #>>43576539 #>>43576610 #
1. skeeter2020 ◴[03 Apr 25 22:50 UTC] No.43576409[source]▶
that might solve your sourcing problem, but now you need to have faith it will draw conclusions and parallels from the material accurately. That seems even harder than the original problem; I'll stick with decent search on quality source material.
replies(1): >>43576567 #
2. somerandomness ◴[03 Apr 25 23:10 UTC] No.43576567[source]▶
The solution is a citation mechanism that points you directly where in the source material it comes from (which is what we tried to build). Easy verification is important for AI to have a net-benefit to productivity IMO.