Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The AI Scientist: Towards Automated Open-Ended Scientific Discovery

sakana.ai

Post reply on HN
Print view View original

31–40 of 144 posts

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#31

Post by malux85 »

I’m working on this now, I literally have another window open beside this browser window with the Multi-agent LLM logs outputs scrolling.

A few differences through - I’m working on Materials Science only. Mine has vision capabilities so it can read graphs in papers. Mine has agentic capabilities too, so can design and then execute simulations on Atomic Tessellator (my startup) by making API calls - this actual design and execution of simulations is what I aimed for at the start.

Long way to go, but there’s a set of heuristics that decide which experiments to attempt which means we only attempt ones more likely to work, lots of fine tuning prompts, self critique, modelling strategies and tactics as node graphs to avoid getting stuck in what I call procedural local minima, and loads more…

I started with MetaGPT framework but found it’s APIs too unstable so I settled on AutoGen, you don’t really “need” a framework, just be sensible about where your abstraction boundaries are, make them simple but composable, Dockerize and k8s for running, and I modified the binaries of a bunch of quantum chemistry software so that multi GPU arches are supported without re compilation (my hardware setup is heterogeneous)

Even if the LLMs can’t innovate in a “new sense” certainly having them reproduce work in simulations for me to inspect is very valuable - I have the ability to “fork” simulations like you can fork code so it’s easy to have the LLMs do a bunch of the work and then I just fork and experiment myself

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#32

Post by foolfoolz »

uniqueuid wrote: post #25
unraveller wrote: post #7

But OpenAI said LLMs can't innovate until human-level reasoning and long-term agenthood is solved. [1] Referring to their precious 5 stages to classify AI before it reaches the scary "beyond" levels of intelligence... presumably at that point they get the feds involved to reg cap the field, so genuine is the fear of the pace they've set. It's clear OpenAI is a hype company knocking over glass bottle stacks at its own…

It would also be sad to see the scientific system destroyed by a wave of automatically generated papers that no human has the capacity to verify. It's not hard to generate ideas, it's hard to generate reliable and relevant ideas. Such AI science generators are destroying the grass they graze on unless they take science more seriously (and not as a toddler idea of "generating and testing ideas", which is only a small…

we already have a wave of papers that no human has the capacity to verify

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#33

Post by surfingdino »

MattJ100 wrote: post #26
surfingdino wrote: post #23

Earlier quoted context omitted.

It will analyse what it is given, but it will not have the ability to say, "hang on, these results are interesting, I wonder what will happed if I pour a different liquid into the drum and spin it at the same speed?" LLMs, especially the latest ones, are decent at analysis of input, but disappointing at producing creative output. I am running a series of experiments using Gemini 1.5 and found it capable of producing…

Interestingly your comment is the very opposite of my experience with LLMs. You can rely on them for creative stuff (write a poem, short story, etc.), but you cannot depend on them for factual stuff. Time and time again they will state "facts" that turn out to be false, and so now I no longer trust it for anything without manual verification. And since I then need to do the research myself for the verification, I rar…

I revisit LLMs a couple of times a year to see if they have gotten any better and it's not a great experience, but I will admit they have gotten slightly better. Still lack and will lack ability to understand what they are processing.

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#34

Post by uniqueuid »

foolfoolz wrote: post #32
uniqueuid wrote: post #25

Earlier quoted context omitted.

It would also be sad to see the scientific system destroyed by a wave of automatically generated papers that no human has the capacity to verify. It's not hard to generate ideas, it's hard to generate reliable and relevant ideas. Such AI science generators are destroying the grass they graze on unless they take science more seriously (and not as a toddler idea of "generating and testing ideas", which is only a small…

we already have a wave of papers that no human has the capacity to verify

Maybe, maybe not. It's a tiered system - you get the deluge at the unfiltered bottom and a narrower selection the more prestigious and selective the outlets / conferences / journals are.

Problem is, of course, that selection criteria are in large parts proxies, not measures of quality. With AI, those proxies become tainted and then you get an explosion of effort.

If anyone has a good recommendation for scalable criteria to assess the quality of papers (beyond fame haha) I'm all ears.

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#35

Post by viraptor »

surfingdino wrote: post #23
OutOfHere wrote: post #15

To produce scientific work, one needs certain raw materials: 1. Data 2. Access to past works Once you have these, only then can discoveries can be made, and papers be written. How does this software get these? I am assuming they have to be provided up-front to the software for each job.

It will analyse what it is given, but it will not have the ability to say, "hang on, these results are interesting, I wonder what will happed if I pour a different liquid into the drum and spin it at the same speed?" LLMs, especially the latest ones, are decent at analysis of input, but disappointing at producing creative output. I am running a series of experiments using Gemini 1.5 and found it capable of producing…

Your experience contradicts the data from the paper. Sonnet generated almost entirely novel concepts, GPT achieved ~4/5 novel ideas. Maybe it's specific to the area of research and the way of prompting, but "it will not have the ability to say" seems to be proven wrong already.

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#36

Post by ben_w »

unraveller wrote: post #7

But OpenAI said LLMs can't innovate until human-level reasoning and long-term agenthood is solved. [1] Referring to their precious 5 stages to classify AI before it reaches the scary "beyond" levels of intelligence... presumably at that point they get the feds involved to reg cap the field, so genuine is the fear of the pace they've set. It's clear OpenAI is a hype company knocking over glass bottle stacks at its own…

> It's clear OpenAI is a hype company

Every other industry: "My new invention is safe, I swear"

Public reaction: "You're biased, it's dangerous!"

Almost the entire AI industry, including people who resign to speak more openly about the risks: "This may kill literally everyone, none of us knows what we're doing or what 'safe' even means"

Public reaction: "You're biased, it's safe, just saying that to look cool!"

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#37

Post by agubelu »

I worked with some people who were actively working on this last year, focusing on CS research.

The biggest issue was validation. We could get a system to spit out possible research directions automatically, but who decides if they're reasonable and/or promising? A human, of course. Moreover, we gave different humans the same set of hypotheses to validate and they came back with wildly different annotations.

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#38

Post by euroderf »

tomohelix wrote: post #10
agf wrote: post #5

> For example, in one run, it edited the code to perform a system call to run itself. This led to the script endlessly calling itself. In another case, its experiments took too long to complete, hitting our timeout limit. Instead of making its code run faster, it simply tried to modify its own code to extend the timeout period. They go on to say that the solution is sandboxing, but still, this feels like burying the…

The beginning of the AI uprising lmao

> it simply tried to modify its own code to extend the timeout period.

And slacking off, at that.

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#39

Post by sampo »

ben_w wrote: post #36
unraveller wrote: post #7

But OpenAI said LLMs can't innovate until human-level reasoning and long-term agenthood is solved. [1] Referring to their precious 5 stages to classify AI before it reaches the scary "beyond" levels of intelligence... presumably at that point they get the feds involved to reg cap the field, so genuine is the fear of the pace they've set. It's clear OpenAI is a hype company knocking over glass bottle stacks at its own…

> It's clear OpenAI is a hype company Every other industry: "My new invention is safe, I swear" Public reaction: "You're biased, it's dangerous!" Almost the entire AI industry, including people who resign to speak more openly about the risks: "This may kill literally everyone, none of us knows what we're doing or what 'safe' even means" Public reaction: "You're biased, it's safe, just saying that to look cool!"

20 to 30 years ago, activists firebombed university research labs (e.g. Michigan State University, University of Washington, Michigan Technological University [1]) because they believed genetically engineered plants are dangerous. Today, we don't have such serious activism against AI. So you are right, the public doesn't think AI is a danger.

[1] https://en.wikipedia.org/wiki/Earth_Liberation_Front#Notable...

Re: The AI Scientist: Towards Automated Open-Ended Scientific Discovery

#40

Post by zipy124 »

As a scientist in academic research, I can only see this as a bad thing. The #1 valued thing in science is trust. At the end of the day (until things change in how we handle research data, code etc...) all papers are based on the reviewers trust in the authors that their data is what they say it is, and the code they submit does what it says it does.

Allowing an AI agent to automate code, data or analysis, necessitates that a human must thoroughly check it for errors. As anyone who has ever written code or a paper knows, this takes as long or longer than the initial creation itself, and only takes longer if you were not the one to write it.

Perhaps I am naive and missing something. I see the paper writing aspect as quite valuable as a draft system (as an assistive tool), but the code/data/analysis part I am heavily sceptical of.

Furthermore this seems like it will merely encourage academic spam, which already wastes valuable time for the volunteer (unpaid) reviewers, editors and chairs time.

Post reply on HN