Forgot your password?
typodupeerror

Comment Automatic verification (Score 2) 84

Some important background for this is that, around 1900, mathematics got formalized to the point that there is a standard for whether a proof is valid or not that doesn't depend on humans thinking about it, just doing a bit calculation following an algorithm. However, it is such a large calculation that nobody wrote complete formal proofs or verified all of the steps (since they'd be spending huge amounts of time on obvious parts, and not the novel aspects). Then, in the past 20 years, people have made computer systems that work on complete formal proofs while showing you the interesting parts, with the rest in macros, and check these proofs automatically. This means that finding proofs has the form of things like what AlphaZero is good at: something to check if moves are legal and a win condition. However, only a relatively small portion of mathematical research has been put into Lean so far, and that doesn't include the sorts of obscure results that often turn out to be useful in other problems. So the current wave is using LLMs to find results in the literature that might be relevant and producing formal versions of them, game AI to put them together, and deterministic validation to check each of these as it goes along and then people can see at the end that the calculation shows the proof is valid. The only part requiring mathematicians is checking that what the proof proves is what they understood the open question to be.

Comment Re:Win 10 support extended... (Score 1) 84

A simplistic view. '"really" don't like' comes down to, in my experience:

- Rejection of change, Doesn't matter what the change is. Just move the Start Menu from left-justified to centered. That's juvenile.

- No more than rumors of enhanced telemetry becoming a serious privacy concern. Rumors are enough. Add in captures of new data and the screamers will blow it up into disproportionate danger. Ignoring their browser, every web site that clandestinely keeps logs that describe your activities, etc. That's a little juvenile also.

- Any launch bugs or flaws. Claimed as proof of rushed release, danger, and well just keep going on.

- Any security concerns, any. As if not merely Widows, but Linux, BSD, even Apple software suffers a security flaw at leas occasionally, some undiscovered for a decade or more.

- And other juvenile complaints and reactions.

Occasionally, Microsoft has hit a release out of the park. W4W 3.11, NT4.0, Windows 8. Mostly because these were preceded by less than excellent releases, though some were leaps ahead that could not be compared to predecessors. Windows 11 is working. Get over it. Or hide behind some excuse and cling to Windows 10 like an infant to their mommy's leg. Linux isn't winning Windows 10 clingers, nor Windows 11 runaways. And if you're paying attention, it's not even winning security and privacy wonks, since they know it takes some effort to secure Linux and hold on to even the illusion of privacy on the modern Internet.

Comment Social Engineering scams so feasible now... (Score 1) 137

I got a 'cold call' from what was actually a legit AI bot just a week ago (from a financial services company serving the loan for some home improvements).

stupid thing never understood "let me talk to a human" no matter how many times i tried.

it was legit, but i just gave up, hung up, called the same number back, and hitting the '0' operator, got a human who handled the situation in a matter of minutes, no frustration involved.

The key thing is the bot cold-called ME but wouldn't provide ANY verifying information that it was what it claimed to be. They have an account # for me, but they wouldn't tell ME it to confirm they were who they were and my account was what they expected to see. Not even something like "last 4 digits".

I mean, they ask US for last 4 digits of something but they won't do the same for us in return?

THIS IS A DAMN PROBLEM, because yeah, we need to have a way to trust that bot-calls from AI are who they say they are, because it is REALLY F'ING EASY to program a bot to do the psychic equiv of a 'cold read' as part of social engineering a scam.

Comment bookworm in progress... (Score 1) 47

I see some of these CVE's are fixed in trixie but not bookworm 6.1.177-1 (yet).

Still other CVE's are fixed in bookworm but not yet in trixie. Seems like a few reboots are in the offing.

First time I've seen several fixes in oldstable before stable. It must be a wild week over there. n.b. fixed, not 'not vulnerable'.

Glad the oldstable crew is hard at work!

Comment Re:Stupid law (Score 1) 44

Is anybody aware of open source hardware/software to do the page turning and photograph/OCR the pages?

Google's rig used a puff of air, IIRC.

I asked somebody in the field and they said, "oh, you want a professional rig - those cost $10K and up.

I was like, "no, I have several rare books I'd like to index and this setup is far less complicated than my $249 3D printer." (I already have a good camera.)

I thought I'd find a hackaday or something but my cursory search wasn't fruitful.

Comment Re:Seems like a job for the OS (Score 1) 28

But how long would it take to implement in every supported OS version and then get all the developers to declare a manifest and update their app?

A policy change can be enforced tomorrow (for better or for worse).

The AI bros might even get Samsung to change their mind. Apparently their 'subcontractors' use this technique to scrape all the content they use for training.

Would threatening to cancel a year's RAM order from the chip subsidiary cause a decision to change in the TV subsidiary? I have no idea how dispersed and decentralized Samsung is.

I'd bet it will stick, though, which is likely for the best. Odds are people aren't running Tor exits on their TV's. Maybe there are legitimate apps I'm not aware of.

Comment Flying Saucers (Score 1) 53

People keep saying it's hard to get a non-blurry picture of them.

Some have suspected somebody has gravity bubble tech which would bend the light.

This thing just spins but isn't itself a disc, though on average it's s disc while in motion.

Post WWII-Germany/Paperclip there were more visible saucers seen over Washington and such but the Germans were developing solid discs. The blurry ones are more over the past few decades as drone tech has become common.

Curious.

Comment Re:PDF (Score 4, Interesting) 77

> I don't know why anyone would send one in any other format.

I'm told by attorneys that Office is nonnegotiable in Law.

Microsoft could probably raise the price to $500/mo and they'd get it.

Should a citizen be required to pay a creepy Redmond company to interact with the Courts?

It"s at least Unconstitutional in New Hampshire, likely other states as well.

Not that Courts themselves care about Unconstitutionality very often. Nor do they personally pay anything for Office. Empathy tends to be low.

But if one wants to make a decent argument that is a good one.

Comment Re:Pedestrians get sent round traffic circles (Score 1) 127

when hiking/walking, i'd rather it stay northbound and click that compass to keep it that way.

otherwise, the pace i'm walking is so slow that the thing keeps changing up what direction it thinks i'm walking in, spins around 3 times, and throws a whole new side trek onto my route.

when i keep it northbound, it still adds that useless side-trek, but at least it isn't spinning all the time.

Comment Re:Rescuer perspective (Score 1) 127

5000 feet (or even 1000) is deceptively tough when it is in the other direction. (I know the Rescuer knows this but saying it for the general).

Canyons and mountain routes can be deceptively difficult. Examples include the Grand Canyon, obviously, but also, say, starting a hike from a high point along the Blue Ridge Parkway or (in Shenandoah NP) Skyline drive. You head down, not up, a generally easy descent to see the landmark...and then after having been out for 2-3 hours, you have to climb back up. So the hardest part of the hike is when you're already tired...

Park rangers and most guide books will remind guests of this, but not every app (esp not google maps) will.

For biking, google maps for desktop is reasonably well at giving you a chance to see the elevation changes on the route to figure out where the toughest climbs are going to be...but I think it still presumes a steady pace rather than the "I quit, I'm getting off and walking" pace that my unconditioned 50-something legs force me to do.

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.”

Comment Re:that's a dumb take (Score -1) 82

> How many people does it take to run a data center after taking up all this property and all this land that could have been used for something else?

Some medium number of H1B's?

No, the jobs are not going to the citizens of the county where the citizens are subsidizing for all this for their own enslavement.

Who would ever believe such a thing after the Covid lies?

Slashdot Top Deals

"Plan to throw one away. You will anyway." - Fred Brooks, "The Mythical Man Month"

Working...