Comment Re:Unencrypted? (Score 1) 128
I've work in technical industries, and they can't handle encryption.
lol wut? DCPs are usually heavily encrypted. Even the dumbest intern running their screening room could figure it out.
Comment Re:Governments don't have the right (Score 1) 124
Governments can, and do, do whatever they want. The guardrails of the "Rule of Law" have been removed. The Constitution is being violated on a daily basis and nothing is being done about it. The Rule of Law is something that is given lip service to while finding every way to ignore it.
Comment Re:Here's the thing (Score 1) 62
At that point we might be looking at the implementation of the Great American Firewall, and the United States would no longer be a representative democracy.
The USA is already NOT a representative democracy. I am uncertain why you think it still is. When Citizens United was decided, representative democracy was condemned. We have seen the results.
Submission + - A fundamental flaw leaves LLMs strikingly vulnerable to attack (technologyreview.com)
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.”
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.”
Comment Re:Trump is RIGHT. Wake up, you Communists (Score 1) 146
Americans are not free. Most are shackled to debt in hope of buying happiness, working jobs they don't want to. They work longer hours, have less holidays and earn less than many countries in the western world. Many Americans are poorly educated, obese and unwell, they suffer a variety of addictions drugs, alcohol, food, social media. They have no public healthcare provision, disregard for the environment and their lifespan is reducing. Half of Americans hate the other half, because they are led by the most selfish, greedy and ruthless who are respected like peasants respect a king, the other half are trying to protect themselves from the other half. Americans have been engaged in some kind of war somewhere since their country's formation and when not fighting overseas they often shoot each other, kids included.
Aside from all of that, USA is still the best. Ra ra ra! USA! USA! USA! (vomit)
You failed to mention the worst part: Rule of Law is a thing of the past.
Comment Re:Sunshine is the best disinfectant (Score 1) 76
Anyone who tells you they can see exactly where the road is heading a mile ahead is selling you something.
I can guarantee you that none of the big players will 'discover' strong AI within the next mile. The money is too infatuated with the current capabilities.
Comment Re:Holy Shit! (Score 1) 49
I'm glad the perpetrators went to prison
The executives received home confinement, NOT prison.
Comment Re:Bullies... (Score 1) 49
And I would argue that, in this case, the punishment was far too lenient, because if you think of the crime in this context, it really says that these people DID get away with it. They got a slap on the wrist. The company paid the bulk of the damages. And so, this settlement again signals that, yes, the privileged class is above the law, because even when the law finds them culpable, the penalty is nominal for them. It's a cost of doing business.
Yes, the punishment was far too lenient; however, it is more serious than you give it credit for. They lost some money directly and they have lost all future earning power except through investments. They will NEVER work anywhere again except maybe as a janitor. Of course, they are generally set for life barring major catastrophe, so yeah, the punishment was WAY to light.
Comment Re:The evidence (Score 1) 49
How are these people not in prison?
They are. The summary states that seven former employees pled guilty to criminal charges. I know it's tradition not to RTFA, but you are supposed to at least RTFS.
Um, the executives mostly got home confinement, not prison. Prison was reserved for the lower level offenders. The only "real" penalties the executives received were fines; with the CEO paying 2 million while the lowest level executive paid $50k. I suspect I read a different article than you did as I kept with tradition and did not read the linked article.
Comment Re:The evidence (Score 1) 49
I wonder how many more campaigns of harassment are ongoing, led by people who know how to conduct vehicle-based surveillance, or encouraged by ineffective or lazy policing.
If you exclude the government and collection agencies, 12,384 currently active. Most are small time (less than $10 million in assets of all parties).
Comment Re:it is destroying evidence like it or not (Score 1) 182
Simply accusing someone of some unspecified crime isn't a right the state has to just bypass Constitutional protections, namely the 4th and 5th Amendments.
Apparently, they do have that right. They did it. Nobody stopped them. Nobody complained about it. They will do it again. They have actually arrested someone for not submitting.
Whether that right is written down or not, it is now established. Rule of Law is whatever whomever is in power wants it to be.
Submission + - AI Companies Destroying Books At Scale (futurism.com)
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.
Submission + - Is Mark Zuckerberg Actually TRYING to Destroy Meta? (futurism.com)
fjo3 writes: Lately, it almost feels like Zuckerberg is trying to destroy his own company. He’s burning through mountains of cash in a desperate attempt to keep up in the AI race. Yet despite the untold billions it’s spent so far, it’s being destroyed by OpenAI and Anthropic. Its employees have even resorted to using AI models made by its competitors, a humiliating reality check for how far behind Zuckerberg’s efforts have fallen.
It’s also a financial nightmare. Capital expenditures have risen dramatically, erasing any appetite for Meta on Wall Street. Its stock price “has been dead money for more than a year,” as Yahoo Finance notes, as investors continue to debate whether Zuckerberg is chasing the AI industry as it careens off a cliff — or edging ever closer to an AI-fueled industrial revolution.
It’s also a financial nightmare. Capital expenditures have risen dramatically, erasing any appetite for Meta on Wall Street. Its stock price “has been dead money for more than a year,” as Yahoo Finance notes, as investors continue to debate whether Zuckerberg is chasing the AI industry as it careens off a cliff — or edging ever closer to an AI-fueled industrial revolution.
Submission + - The first active shooter suppression service
An anonymous reader writes: Schools to tackle active shooters with pepper-spraying drones that can ram attackers
Pepper-spraying drones aimed at tackling active school shooters are being rolled out at campuses across three states.
The drones, part of the Campus Guardian Angel program, can ram into targets at speeds of up to 70 mph and are capable of weaving through the air to avoid gunfire. Three schools in Florida, five in Georgia and one in Colorado are part of pilot schemes using the aerial technology.
Pepper-spraying drones aimed at tackling active school shooters are being rolled out at campuses across three states.
The drones, part of the Campus Guardian Angel program, can ram into targets at speeds of up to 70 mph and are capable of weaving through the air to avoid gunfire. Three schools in Florida, five in Georgia and one in Colorado are part of pilot schemes using the aerial technology.
Submission + - Your Brain Can Rewire Itself to Allow True Multitasking (sciencealert.com)
alternative_right writes: Our daily lives are built on multitasking, but are our brains actually doing two things at once, or just switching very quickly between them?
Well, it depends. The science shows it's the latter, for cognitively demanding work.
But in a new study published in the Journal of Cognitive Neuroscience, researchers from Georgetown University Medical Center in the US have revealed that we can put certain tasks on autopilot in a way that enables something closer to true multitasking.
Driving is the perfect example: When you take your test, your entire mental and physical energy is concentrated on executing the right combinations of movements and thoughts.
After a decade behind the wheel, the brain is no longer consciously thinking everything through in great detail.
Well, it depends. The science shows it's the latter, for cognitively demanding work.
But in a new study published in the Journal of Cognitive Neuroscience, researchers from Georgetown University Medical Center in the US have revealed that we can put certain tasks on autopilot in a way that enables something closer to true multitasking.
Driving is the perfect example: When you take your test, your entire mental and physical energy is concentrated on executing the right combinations of movements and thoughts.
After a decade behind the wheel, the brain is no longer consciously thinking everything through in great detail.
