AI cracks superbug problem in two days that took scientists years

News already corrected.<p>&quot;Google Co-Scientist AI cracks superbug problem in two days! — because it had been fed the team’s previous paper with the answer in it&quot; <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=43162582#43163722">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=43162582#43163722</a><p>Let&#x27;s see how many points gets the correction. It would be good that achieved the same or more visibility than this one to keep HN informative and truthful.

It says it took them a decade, but they obviously published loads of intermediate results, as did anyone else working in the space.<p>I get that they asked it about a new result they hadn&#x27;t published yet, but the idea that it did it in two days when it took them a decade -- even though it&#x27;s been trained on everything published in that decade, including whatever intermediate results they published -- probably makes this claim just as absurd as it sounds.

&gt; <i>Critically, this hypothesis was unique to the research team and had not been published anywhere else. Nobody in the team had shared their findings</i><p>This seems like the most important detail, but it also seems impossible to verify if this was actually the case. What are the chances that this AI spat out a totally unique hypothesis that has absolutely no corollaries in the training data, that also happens to be the pet hypothesis of this particular research team?<p>I&#x27;m open to being convinced, but I&#x27;m skeptical.

&gt;He gave &quot;co-scientist&quot; - a tool made by Google - a short prompt asking it about the core problem he had been investigating and it reached the same conclusion in 48 hours.<p>Could it be the case when asking the right question is the key? When you know the solution already it&#x27;s actually very easy to accidentally include some hints in your phrasing of question that will make task 10x easier

Using this, launched yesterday:<p>Today Google is launching an AI co-scientist, a new AI system built on Gemini 2.0 designed to aid scientists in creating novel hypotheses and research plans. Researchers can specify a research goal — for example, to better understand the spread of a disease-causing microbe — using natural language, and the AI co-scientist will propose testable hypotheses, along with a summary of relevant published literature and a possible experimental approach.<p><a href="https:&#x2F;&#x2F;blog.google&#x2F;feed&#x2F;google-research-ai-co-scientist&#x2F;" rel="nofollow">https:&#x2F;&#x2F;blog.google&#x2F;feed&#x2F;google-research-ai-co-scientist&#x2F;</a>

So the AI didn&#x27;t prove anything, it offered a hypothesis that wasn&#x27;t in the published literature, which happened to match what they&#x27;d spent years trying to verify. I can see how that would look impressive, and he says that if he&#x27;d had this hypothesis to start with, it would have saved those years.<p>Without those years spent working the problem, would he have recognized that hypothesis as a valuable road to go down? And wouldn&#x27;t the years of verifying it still remain?

Here is the scientists Google scholar account. He has been publishing (and so have lots of other scientists, just look at the papers HE cites in previous work) about this exact scenario for years. This is just Google announcing a brand new AI tool (the scientist tool was literally just released in the past few days). And knew mainstream outlets have 1. No clue about how AI works, and 2. A pathological deference to anyone with lots of letters after their name from impressive places.<p><a href="https:&#x2F;&#x2F;scholar.google.com&#x2F;citations?hl=en&amp;user=rXUHiP8AAAAJ&amp;view_op=list_works&amp;sortby=pubdate" rel="nofollow">https:&#x2F;&#x2F;scholar.google.com&#x2F;citations?hl=en&amp;user=rXUHiP8AAAAJ...</a>

A thought occurred to me, as someone involved in some projects trying to recalibrate invectives for science funding: Good grad students are usually more intersectional across fields (compared to supervisors) and just more receptive to outsider ideas. They unofficially provide a lot of this same value that AI is about to provide established researchers.<p>I wonder how AI is going to mess with the calculus of employing grad studies, and if this will affect the pipeline of future senior researchers...

When I read &quot;cracks superbug problem&quot;, I thought AI solved how to kill superbugs. From reading the article, it seems like AI suggested a few hypotheses and one of which is similar to what the researcher thought of. So in a way, it hasn&#x27;t cracked the problem although it helped in forming ONE of the hypotheses, which needs to be tested in experiments(?)<p>Just want to make sure I&#x27;m understanding what&#x27;s written in the article accurately.

why did it take 48 hours? did he give the AI data to process or was it just a prompt. did it spit back out a conclusion or a list of possible scenarios it had scraped? seems like a PR stunt.

&quot;AI&quot; didn&#x27;t &quot;crack&quot; anything here. An LLM generated text that&#x27;s notionally similar to a hypothesis this researcher was interested in but hadn&#x27;t published. You can read Dr. Penades in his own words on BioRxiv, and if you might have been interested in reading the prompt or the output generated by co-scientist, it&#x27;s included in the results and SI: <a href="https:&#x2F;&#x2F;www.biorxiv.org&#x2F;content&#x2F;10.1101&#x2F;2025.02.19.639094v1" rel="nofollow">https:&#x2F;&#x2F;www.biorxiv.org&#x2F;content&#x2F;10.1101&#x2F;2025.02.19.639094v1</a><p>What actually happened here looks more like rubber-ducking. If you look at the prompt (Supplementary Information 1), the authors provide the LLM with a carefully posed question and all the context it needed to connect the dots to generate the hypothesis. The output (Supplementary Information 2) even states outright what information in the prompt led it to the conclusion:<p>&quot;Many of the hypotheses you listed in your prompt point precisely to this direction. These include, but are not limited to, the adaptable tail-docking hypothesis, proximal tail recognition, universal docking, modular tail adaptation, tail-tunneling complex, promiscuous tail hypothesis, and many more. They collectively underscore the importance of investigating capsid-tail interactions and provide a variety of testable predictions. In addition, our own preliminary data indicate that cf-PICI capsids can indeed interact with tails from multiple phage types, providing further impetus for this research direction.&quot;

I find this odd because that&#x27;s exactly how I thought viruses worked when crossing species and I have no background that would lead me to that conclusion and have almost nothing in my life that would make me ponder such a thing.<p>I feel like someone explained this in the 80s to me.

this might be a dupe and the title is completely misleading, it (the AI) simply provided one of the hypothesis (which took two years to confirm) as the top result. A group of humans can come up with these in seconds given expertise in the subject.

Maybe this is too tin-hatty, but sending the unpublished manuscript from (or to) a free Gmail account gives Google the right to use it, and therefore also to train whatever AI model they want with it.<p>Sounds like all of those claims where ChatGPT allegedly coded a flappy bird clone from scratch. Only it didn&#x27;t, it just regurgitated code from several Github repos.

Hey guys, I cracked the general theory of relativity in 30 seconds. Just got to find the right download!

But they needed to have already solved the problem in order to be able to verify that the LLM hadn&#x27;t just hallucinated a solution.<p>(and I wonder how many hallucinated solutions the LLM came up and were rejected - sorry &quot;refined&quot; - by the team).

I&#x27;m getting the impression that this worked becaused the LLM had hoovered up all the previous research on this topic and found a reasonable string of words that could be a hypothesis based on what it found?<p>I think we are starting to get to the root of the utility of LLMs as a technology. They are the next generation of search engines.<p>But it makes me wonder, if we had thrown more resources towards using &quot;traditional&quot; search techniques on scientific papers, if we could have gotten here without gigawatts of GPU work spent on it, and a few years earlier?

I’d like to know how we got an email back from Google confirming that they don’t have access to his computer

There is a real societal danger in ascribing abilities to &quot;AI&quot; that it just doesn&#x27;t have.

&quot;LLMs are great at solving problems that already have solutions.&quot;<p>This is performance art, right?

can&#x27;t wait AI to discover tons of Ramanujan style math inventions.

Lmao. Not at all surprised the BBC didn&#x27;t even bother to look at this guy&#x27;s Google scholar account. He has been publishing papers about this exact scenario for years. So have many other scientists.<p>A few years back I was so sick of blockchain vaporware, and honestly couldn&#x27;t think of anything more annoying... But several years of reading &quot;serious&quot; outlets publish stuff like &quot;AI proves the existence of God&quot;, or &quot;AI solves cold fusion in 15 minutes, running on a canon R6 camera&quot; makes me wish for the carefree days of idiots saying &quot;So you&#x27;ve heard of Uber, now imagine Uber, but it&#x27;s on the blockchain, costs 0.1ETH just to book a ride, and your home address is publicly accessible&quot;...

The article is pretty fucking trash, tbh.<p>The &#x27;&#x27;&#x27;journalist&#x27;&#x27;&#x27; is, too:<p>Author: <a href="https:&#x2F;&#x2F;muckrack.com&#x2F;tom-gerken&#x2F;articles" rel="nofollow">https:&#x2F;&#x2F;muckrack.com&#x2F;tom-gerken&#x2F;articles</a>