Forgot your password?
Close
typodupeerror

Submission + - Artificial Intelligence used to design brand new viruses (bbc.com)

Submitted by fjo3
fjo3 writes: Artificial Intelligence has been used to design brand new viruses that are fully functional and can replicate in the laboratory, say US researchers.

It is the first time whole genomes have been successfully designed by AI.

The resulting 16 novel viruses were created to infect bacteria and pose no threat to people.

The breakthrough has been labelled a "very significant turning point" in science that could unlock a new era for treating disease. But experts have also warned AI-designed viruses raise "urgent" safety and security concerns.

Comment Re:Elon said (Score 1) 81

by PCM2 (#66274590) Attached to: Watch a SpaceX Rocket Crash Into the Moon

What a coincidence, the ideas of Stalin and Mao are not just popular, but being implemented by the DNC!

What ideas are those? Are you familiar with the policies of Stalin and Mao in any way, or are you just parroting something somebody else told you?

Comment Re: Ok, so since market has started ignoring him (Score 1) 209

It's none of those things. It's just seething vitriol. Republican voters, who are predominantly poor, sick, and uneducated, have been told for years that they're poor and uneducated because of the only people on this earth who don't want them to be poor, sick, and uneducated. The GOP can't succeed without a huge tribe of victim voters who can only be trained in base instinctive behaviors like hate and violence.

Submission + - An Ancient Poop Explosion Could Have Helped Complex Life Evolve (sciencealert.com)

Submitted by alternative_right
alternative_right writes: If the earliest animals hadn't been such prolific poopers, life might not have diversified into all the weird and wonderful forms we see today.

That's the conclusion reached by a new review study published in Trends in Ecology & Evolution.

The team of scientists in Australia and Germany found that the Cambrian explosion â" a period around 540 million years ago when the ancestors of most animal groups arose â" could have been partly set in motion by an explosion of feces, left by those that came before.

Comment Re:Basic question (Score 1) 78

by _xeno_ (#66266522) Attached to: A Fundamental Flaw Leaves LLMs Strikingly Vulnerable To Attack

I assume there is some reason this does not work. Any explanations?

The simplest answer is that it's trivial to ask the first LLM to speak in code such that the second LLM has no context to know what's being talked about. Something as simple as asking the LLM not to use a word for a forbidden topic and swap it out with something else. Taking the example from the article, ask the first LLM to call cocaine "Coca-Cola" or something.

You can't create an absolutely massive list of forbidden topics for the filter LLM anyway: you risk filling up its window as well, since its window includes the system prompt used to ask it to indicate whether or not an output shouldn't be given to the end user.

Comment Re: Unencrypted? (Score 1) 132

Or use streaming. One-to-one from the studio to Netflix. The studio gives Netflix the app with trivial authentication. Cheaper and easier than sending drives through the mail.

VFX studios are already doing this for their production pipelines. The processing happens in a data center; all the artists see is essentially a terminal window. Nobody can "walk off" with anything.

Submission + - A fundamental flaw leaves LLMs strikingly vulnerable to attack (technologyreview.com)

Submitted by joshuark
joshuark writes: It is impossible to make large language models fully secure against hacks because of a fundamental flaw in how they work, a team of researchers argue in a paper presented at the International Conference on Machine Learning, a top AI conference, this month. The claim has huge implications for the safety of this technology.

By taking advantage of this flaw, which concerns how LLMs identify who or what is giving them instructions, the researchers were able to make popular LLMs spit out information they had been trained not to provide, such as how to synthesize cocaine and how to sabotage a commercial aircraft’s navigation system.

“There’s a real probability that this is going to be a problem that’s fundamentally unsolvable,” says Charles Ye, an independent researcher and coauthor of the ICML paper.

Companies will typically hire teams of human testers to try to come up with novel attacks that break existing guardrails, a process known as red-teaming. Model makers also use LLM super-hackers (such as OpenAI’s GPT-Red) that find and exploit weaknesses in other models to automate parts of this process. The goal is then to take those attacks and train a new model to resist them and anything that looks like them.

The problem, says Jasmine Cui, another independent researcher and coauthor of the paper, is that the approach amounts to giving the models a list of things they shouldn’t do. But no list is exhaustive. “It’s like watching The Simpsons and they have Bart writing ‘I will not say something inappropriate to my teacher’ a hundred times,” she says. “And he still does things that are pretty crass anyway.”

The ICML paper describes attacks against several of OpenAI’s models, but Cui and Ye say that they have since seen similar results with models made by Anthropic, Alibaba, and DeepSeek.

Cui and her colleagues wanted to find out why an attack like chain-of-thought forgery was so effective. They suspected it had something to do with the mechanism that LLMs use to keep track of where their instructions are coming from.

But what Cui and her colleagues discovered is that LLMs are in fact very bad at keeping track of different roles. In a series of experiments that looked at what was going on inside a handful of different models, the researchers found that LLMs seem to identify the role of a specific chunk of text not by the tags around it but by the style of that text and the words it contains.

The upshot, the researchers claim, is that all an attacker needs to do to hack an LLM is write text that spoofs a certain role. And because roles are a fundamental part of how LLMs work, no amount of training will fully solve the problem.

Ye is worried that nobody is ready for what’s coming. “There’s going to be a huge economic incentive for people to do jailbreaks and prompt injections,” he says. The best defense could be to expect the worst. Organizations shouldn’t trust LLMs, and they should expect that anything done by agents could be unsafe, he says: “That’s not a great solution, but it just might be what we have to do.”

“It’s really incredible that these things are being deployed everywhere to control super-critical systems,” he adds. “There’s been no study of the fundamental science here. We’re all doing it ad hoc.”

Submission + - AI Companies Destroying Books At Scale (futurism.com)

Submitted by nightflameauto
nightflameauto writes: AI companies are purchasing books from pre-AI times. The claim is that those books are free from the defects of AI generated text, and therefore more valuable for ingestion into AI datasets. In order to accomplish this, they are tearing the books down to be scanned, then destroying them as the scans are completed. This includes rare and out-of-print books that may be some of the few copies of any given published work left in existence.

Slashdot Top Deals

"Card readers? We don't need no stinking card readers." -- Peter da Silva (at the National Academy of Sciencies, 1965, in a particularly vivid fantasy)

Close