I’m guessing it created it the same way i created a giant eight limbed creature called an Octophant, using a piece of paper and a pencil.
It does not make it exist.
According to this non-paywalled article, it created thousands of potential genomes from which the scientists running the experiment chose three hundred promising candidates. They then had bacteria make those and used them for their bacteriophagy experiment. So essentially the AI made the blueprints that the scientists then turned into real stuff.
Let’s get real here. Could the bacteria make my Octophant if I clean up the pencil lines?!
Cat girls first dangit! Get in line.
Unfortunately the science hasn’t gotten that far yet, but maybe in the future.
So the AI’s role in all this, given that creating new viruses in a lab is already understood and well-practiced, was just to generate a list of virus configuration options (presumably working from existing examples), from which scientists had to discard ~90% of what the AI generated before moving on to the next step.
Sounds like an attempt at a headline more than a genuine scientific effort.
So the AI’s role in this process, which is already known and well-practiced,
Uh… no? Creating custom-made genomes for curing specific diseases isn’t “known and well-practiced;” it’s an area of active research.
from which scientists had to discard ~90% of what the AI generated before moving on to the next step.
That’s basically in line with how AI is used in similar applications; even if most of what you get is busts AI can make so much stuff that it can compete with (or even outdo) other alternatives by sheer quantity. Sure someone has to filter out the useful stuff from the trash, but the very fact that it produces useful stuff makes it worth it. This method is well-established in other areas, so there’s no reason to dismiss it here.
Scientists had no clue at all how to design bacteriophages is my understanding. Only with the help of LLM have they been able to generate the first man-made bacteriophages ever.
Same question to you, why would you do this?
Consider the military and industrial applications of the octophant.
[You Have Been Reported to the Raytheon Ethics Board]
[All Clear. Proceed]
[Congratulations on your Sydney gottlieb ethics in STEM award! The bonus should show up in your account by Monday!]
Oh boy a promotion!
Same answer: The hubris of man.
Same way we all created the sun, and built a house next to it as children. All without picking up a hammer. Lazy kids.
A version of this came out of a DND campaign. They’re called octen. Like oxen, but octen. They’re like regular oxen, except they have eight legs. The legs have the same fur and color as cattle legs, but then bend up and move like a tarantula’s.
Yes and no. Going from a design to production takes remarkably little time these days.
So you just scan his picture of an octophant and print it out and it exists? I’d bet you have a bunch of meat and no life. That said, please spend the time and effort into something more productive instead of a octophant was the point
That said, please spend the time and effort into something more productive
Like curing antibiotic-resistant bacterial diseases? Because that’s what the experiment was for.
An octophant is a bacterial disease? Lol
And yes this experiment may be an absolutely stupid thing to do. Could it work out in a extremely well regulated extremely tight secure facility, sure.
Is that happening, not in the U.S., we are cutting all the regulations and health precautions, so if what they create ever kills good bacteria as well during the tests, you should assume it will be spreading to the population.
Why wasn’t it contained? Our research grants were cut in half so we skipped on the precautions. Also the regulatory board no longer exists, and our OSHA equivalent hasn’t stopped by in over 2 years. I think they may have shut down
Mine was called Strandaglend and added tempera to the medium. Sometimes I forget the middle name of relatives but never of my creation 42 years ago.
Yes and no. Going from a design to production takes remarkably little time these days.
Are there commercial virus-printing services? If so, what keeps some yahoo from sending them the virus-equivalent of a .stl file for smallpox? Can someone just order virulent pathogens printed on command?
Assuming that it will work from the first try
I mean, it’s a known process.
There’s a YouTube channel called Thought Emporium.
The guy used to do live streams were he designed custom DNA, and then ordered it from a lab that could produce it.
It was a simple commercial transaction.
I used to work in a wet lab. Agree, designing or multiplying DNA is not a problem. The problem is to make it stable and do what you intend it to do. This could be done only by allowing ai to control the whole process with minimal human interaction
Why go for the cures right?
That’s literally what a bacteriophage is.
For the uninitiated, the scientists created bacteriophages, viruses that eat bacteria and are being studied as a way to cure bacterial diseases without using antibiotics. Non-paywalled article: https://www.theguardian.com/science/2026/aug/06/safety-fears-as-scientists-make-first-viruses-designed-by-ai
For real. I hate AI but people need to read the damn article.
Sorry, but the anti-AI movement is about as well informed as the anti-vax and flat-earth movement.
I am pro-AI and yet fully agree it needs to be ethically and responsibly used and regulated. Data super-centers bad, Grok bad, ChatGPT bad, etc.
I have even seen people say “LLMs can only be chatbots and all chatbots are bad!” Not understanding that this type of research, as well as weather prediction, and animal behaviour studies, all use LLMs.
And every time I’ve attempted to broach the subject, I’ve been accused of being an OpenAI fanboy, etc.
The “anti-AI movement” is just most people on earth, so of course there’s millions of stupid people them.
The pro AI people are the odd ones out, like my boss who vibe coded a task planner and then showed it to us and told us to use AI to develop our complex app faster, as if it isn’t built into every IDE nowadays and he is some genius who just needs to tell the stupid nerds how things work. Just a reeeally special bunch of people.
So you advocate for using AI to develop new drugs? Study weather patterns? Assist in physics and material sciences? Assist with coding? Are you saying pro-AI people think all AI should be embraced at every level, and replace humans?
Did you rely on AI to help you understand my comment?
The point is that calling it an “anti-AI movement”, is like saying there’s an anti shit eating movement. Yes, there are some weirdos being an exception, but apart from that it’s just humanity as a whole.
And I reject that premise entirely, which is why I asked directly whether or not you supported the usage of AI I listed. Because from what I have seen, people who reject AI do so wholesale, with no regard to how it is being used. That’s what I mean by the Anti-AI movement, not that you have a club and meetings.
I am not defending your bosses use of AI to vibecode, but that has absolutely nothing to do with what OP posted.
So are you condemning an entire technology and field of research on the grounds of the worst examples, or do you support the usage of AI in the fields I mentioned?
X is what I mean by Y, not Y
Since we’re rejecting premises, logic rejects this premise. And I reject the premise of me guiding you to a place where it makes sense. Just talk to Grok instead, it will give you the positive affirmation you seek.
pro-ai people are either already delusional, or they are corporate executives which categorically position them to be a priori delusional to ever land in such a job position.
Pro-AI folks are useless in the thinking department, so of course they want to force it into everything - to lay off folks, to cut costs on security and quality, and to generate short-term profits and then fuck off once the bubble inevitably buckles under the weight of reality.
Okay, then you don’t actually understand my position at all. Would you rather tell me what I believe, and reinforce my assertion that you’re misinformed. Or would you like to actually understand my position, and destroy the strawman that you have built?
Agreed. Not only uninformed but also irrational.
So what? I eat bacteria all the time.
The title is misleading. It wasn’t an LLM, but a bespoke, dedicated system.
And yes, I know that’s sort of “AI,” but these days when that word is used, the average person equates it to ChatGPT or Gemini or whatever text chatbot they’re familiar with.
This basically has nothing to do with that. Scientists have been making models for design and data synthesis forever.
It was an LLM trained on mountains of research.
I don’t think so. I don’t have access to the research, but almost assuredly it is a custom neural network for specifically things like this. An LLM would never be able to describe in detail the virus to a point that researchers would be able to fabricate it. That’s just not how that works.
You don’t understand. This isn’t a debate. They used Evo1 and Evo2 which are both LLMs. That is just a fact. You don’t understand what LLMs are. I’m not trying to be mean or rude, it’s simply true if you think an LLM couldn’t do this, because it literally did.
Edit: sorry, that probably came of extremely rude and short tempered, and I shouldn’t treat you that way.
LLMs are not only chat-bots, and in fact so not even need to “speak” any language you or I would recognize. Evo 1 & 2 are LLMs built off of raw genomic data. Literally the raw A, C, G, T pattern of the genetic code itself. Trillions of nucleotides. So the output of the LLM wasn’t “Here’s what a bacteriophage would look like!” It was “AT,CG,AG…” Raw genetic code which can then be synthesized.
Okay so, am I crazy here or does that mean it’s just a machine learning neural network. Which is what LLMs are, but LLMs are specifically designed for language. That’s their whole purpose. If this was not designed for language isn’t it by definition not a LLM.
Yes, language – a LLM trained only on English writing will have zero ability to respond to a question in Spanish. These LLM are trained in the language of genetics (e.g. ACTG).
It’s like putting an internal combustion engine in a car or a boat. One navigates land and the other water, but the engine can actually just be the same one.
You’re mixing up colloquial usage with scientific usage. Scientists are notoriously bad at giving things names (thanks dark matter…) All LLMs are neural networks. All neural networks are AI. Not all AI are neural networks. Not all neural networks are LLMs. This is a case of nested hierarchies, like phylogenetic taxonomies. You are a bony fish.
So yes, LLMs are literally just a form of neural networks, but the language they learn on does not have to specifically be English, or German. And this “Language” was genetic code. So Evo1 and Evo2 Literally speak genetic code. That is their language, since it is what they were trained on. They are predictive text models that can predict what the next genetic sequence will/should be.
That thing looks like it’s a 'billion" residues long. Granted I have only rudimentary understanding of proteins, as I did my master years ago on thermodynamics, but using enzymes as the system.
How can you be sure anything in there will fold correctly. That you can design and describe a synthetic virus that in theory will work, but actually making it?
Edit: Ah, normal GMO, just used the models to tell them what to modify. Not a complete synthetic virus. Probably just did some normal injection of new lines in existing viruses.
Technically, not misleading. It’s that LLMs are not AI. The world is misleading! Title is actuate!
AI is a massive umbrella, under which LLMs are categorized.
That’s objectively false. LLMs are neural networks, which is a form of AI.
Don’t fucking post images of headlines.
Find a source that isn’t paywalled and link to the article. Or don’t post anything at all.
Or fuck off? Like is it harder to read in a picture? Want me to clean your screen, wipe your ass too?
People who aren’t illiterate actually read articles. You post the information about the article so that people know if they want to read it or not. Then we discuss the contents of the article with other assholes who also read the article.
Because discussing with others who have also only read the blurb is fucking useless because nobody knows a fucking thing.
Posting the article link both makes it easy to facilitate this while also maintaining the polite fiction that the other dumbasses could have at least in theory read the article.
You are posting here but you have no fucking idea what the fucking point of such forums is because you are an illiterate child and then you have the nerve to get angry. You don’t have to provide a link… you don’t have to make a post either you could leave that task to the actual adults.
Also the culture of posting such pictures results in trivial fakes blending in because idiots judge the veracity of the source pictured without any way to authenticate unlike a fucking url.
🥇gold
Please have my child
Pictures are not reliable sources for information. It promotes a culture of disinformation.
And blind people can’t read words in images.
There, a link to the source of the picture… https://bsky.app/profile/financialtimes.com/post/3msgnagdoh72l you can now bathe in more pixels
Site doesn’t load
Yeah fuck people who use screen readers and/or are broke hahaha what losers.
/s in case it’s needed
Nah, you dont even whip the shit off your own chin after eating all that
“Grok. Engineer me a highly infectious virus. Optimize it to have high lethality to ethnic groups other than my own, but low lethality to my own ethnic group.”
This world is cursed.
It’s the plot of utopia.
“My AI created a virus that cures all autoimmune diseases!”
*squint* “this is just HIV”
AI is a synthetic virus.
This is like watching Contagion from 2011 during the pandemic.
That graphic looks like the cover of an O’Reilly book
The first physical computer virus
See? People still want physical media!
Umbrella Corporation is coming along nicely, I guess?
deleted by creator
Literally my first thought.
You realize your body is host to literally trillions of bacteriaphages right now? “Virus” doesn’t automatically mean universal infecting agent. They need to be calibrated to specific bio-markers. Most bacteriaphages only infect a SINGLE species of bacteria.
And before you mention mutations and cross-species jumps, there is a very big difference between the viruses that infect you and I, and bacteriaphages. There has never, ever, not even one single solitary time, been an instance of a bacteriaphage suddenly mutating to start infecting macro-organisms.
There has never, ever, not even one single solitary time, been an instance of a bacteriaphage suddenly mutating to start infecting macro-organisms.
True and correct. And even modifying towards that isn’t really worthwhile because there’s better delivery mechanisms than bacteriophages for those avenues.
But that doesn’t mean improperly tailored bacteriophagy therapy isn’t a concern for some macro-organisms. Like the kind that tends to operate with biology that is tightly interwoven with microbial life, even to degrees we have not yet fully made in-roads into enough understanding yet. Macro-organisms, like Humans.
They need to be calibrated to specific bio-markers. Most bacteriaphages only infect a SINGLE species of bacteria.
Yes. A real limitation of the approach some might say, especially anything naturally selected for. Wait, did I say might say?
https://pmc.ncbi.nlm.nih.gov/articles/PMC11876824/
Even from the abstract dated two years ago, you should be able to glimpse the fact that there’s multiple approaches under study to circumvent these limits. It is very required if the technology is to live to it’s potential. From nano-particles supplementary structures (with as of yet unknown immune responses), to phage cocktails. A lot of good science that could be easily carried out wrongly in the future, given the right causes or incentives or plain bad luck.
Imagine we get to the vibecoding equivalent of this (I know it seems absurd, but so has everything else)? Epidemics of lactose intolerance, nutrient malabsorption, and mental health crises, because a future FDA went even more…whatever the hell they currently are doing. I want my body’s microbiota to only be subject to my abuses.
Okay, those are fair concerns, and it’s refreshing to see actual founded rebuttals instead of the slack jawed “But science scary!”
True, an improperly built phage could cause some harm. But the goal is to make killers, not modifiers. And most of what you covered is powered by enzymes your body makes, not bacteria in your gut. And worst case scenario, let’s say it does affect some of the microbes that assist in nutrient absorption, the answer to that is to nuke your gut, and get a faecal transplant. Not ideal, but very common.
Could this go awry? Yes, in the ways ALL medical research has the potential to go awry. But this is an objectively positive discovery that is undoubtedly going to save millions, maybe even billions of lives. Antibiotics is something that truly changed the course of our species, but it’s effectiveness is limited, and we’re nearing the end. This is a proposed answer to it, that will prevent you from dying of simple infections.
Demonizing it helps nobody.
Are you a bacterium?
No?
You’re fine
SO WHAT IF I AM?! ARE YOU SAYING BACTERIA-AMERICANS ARE TOO STUPID TO USE THE INTERNET?!
AI SOLVES THE FERMI PARADOX
I wouldn’t say “solves” so much as answers whether our great filter is ahead of us or behind us (it’s the former, btw).
Humans have designed viruses before so ya nothing fermi breaking aqui.
IMO it’s more of a case that there are a ton of great filters, some of which we’ve passed through already, others are still ahead. But all the ones that we’ve previously passed probably mean that intelligent life capable of advanced technology is very rare, possibly even to the point that we’re the only ones in the galaxy.
Bacteriaphages are no more dangerous to complex life than a knife is to single-cellular life.
What? A bacteriophage in your gut could definitely kill you.
bacteriophages only infect bacteria, hence the name. they cant infect eukaryotes, or even things like archaea. also they target specific strains of bacteria, and only the ones that arnt resistant to bacteriaphages.
How? Do you know how they work? What they are? You have literally trillions in your gut right now. There are literally more bacteriaphages inside of you than human cells and bacteria combined.
My thinking was that a novel bacteriophage could potentially clear out the bacteria in your intestines. Seems like natural bacteriophages are all pretty cool with humans, though. https://en.wikipedia.org/wiki/Bacteriophage
And that phage wouldn’t kill all your gut flora, it would only kill one strain/species at most. Which is just a normal Tuesday for your gut microbiome.
Ah, yes, and that’s why you were asking me how they worked, because I clearly did not. Thanks for the education! I wasn’t expecting to learn about bacteriophages when I woke up this morning.
And thank you for taking the time to fill in your knowledge gaps! I really appreciate the willingness and patience.
it wont, each phage, is only able to infect one species, and even some strains only. theres always evolution arms race between bacteria and viruses, so theres is no virus attacks all bacteria. since more than one species exist in your gut. also it gets complicated if you consider bacteria can become resistant to phage infection, yes it can be overcome.
Well, this IS an important milestone towards curea for many illnesses and problems, and yes, this is a field where AI could actually accel and be a great tool
I have a long list of issues with AI, but this ain’t on it.
I guess you mean excel, not accel
Uh, yeah - ‘cause mega ebola will be fun.
Ebola isn’t a bacteriophage. Do you understand what this research actually does? I’m not trying to be rude, but fear about this can only be spawned from not understanding what a bacteriophage is, what it does, and how it does it.
I mean fire was one of the first researched technologies and to this day man made fires kill people – but still I think we agree learning how to make fire has been a net positive for humanity.
Well, maybe not, but what I do believe that if fire is a net benefit for society, then in a similar fashion future technologies have and will be. People will always use the best technology available to them to do evil things – but that’s just the other side of the coin of using the best technology available to do good things.
Maybe use AI to fix your dreadful grammar.
We really are the dumbest of species, aren’t we?
Are you saying this because researchers made a bacteriaphage? Are you under the assumption that they are, or can be, dangerous to us?
On their own? No of course not. But since when have things that can be weaponized not been weaponized?
Because the only way you could weaponize a bacteriophage is by deciding you want to kill one specific bacteria strain in a person.
You could make a bacteriophage to target my favourite brewers yeast, but selecting a single strain of ecoli in my gut wouldn’t harm me any more than eating a giant bowl of extra-extra-spicy salsa.
They don’t have the same mechanisms as other viruses. They physically cannot harm eukaryotes.
Again. At this interval, they’re not a threat per se. But this development isn’t going to stop at this interval.
So you understand what a bacteriophage is? You’re effectively arguing that they are going to weaponize string cheese just because “That’s what they do.” The entire premise is nonsense, unless you have no idea what a bacteriophage is, or how it works.
yup yup yuppers
