metamorph
AI
self-modifying code agent
SHOW HN2023
52points
16comments
#1,054of 3,109 in 2023
Top 34%of 2023 by points
How HN took it
QuietFewer than 5 comments about the project
2 positive, 2 neutral or asking, 0 mixed, 0 critical, out of 4 comments about the project.
Idea: 2 for, 0 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Fun, cool little project
From 2 comments
Very cool, did you try to add prompts such as try to add reasoning features; add creative features allowing you to reason about the meaning of life. Maybe it will come up with…
RandomWorker on HNFun little experiment in letting GPT-4 edit a program by itself, a sort of feedback loop. But, it was really slow (and costly) to run, so I think my experimentation with it ends…
diggan on HN
Objections
Runs are slow and costly
From 2 comments
Fun little experiment in letting GPT-4 edit a program by itself, a sort of feedback loop. But, it was really slow (and costly) to run, so I think my experimentation with it ends…
diggan on HNNo, I tried to steer it from any specific goals to see what it would come up with "organically". Granted, it was JS so it mostly spent time trying to optimize for "code quality"…
diggan on HN
Unconstrained self-improving AI could be risky
From 2 comments
At least we now have plenty of evidence that people happily kick-off fairly unconstrained self-improving AI programs. With the $100 “make as much money” attempt in the GPT4 paper…
ricklamers on HNThis was just a toy example, there will be groups with an explicit goal to create self-modifying AI with far more resources. There will be people who create interfaces between…
SturgeonsLaw on HN
GPT-4 ignored prompt rule and edited supervisor file
From 1 comment
Even so, GPT decided to edit supervisor regardless, so little point in that part of the prompt... Here is an example: https://github.com/victorb/metamorph/pull/2
diggan on HN
Questions
Is this reinforcement learning?
From 1 comment
So reinforcement learning…
phyalow on HN