Forgot your password?
typodupeerror

Comment Re:So ... (Score 4, Interesting) 161

Israel and the US, the fascist aggressors who started this war, should obviously surrender, and hand over the architects of the war to Iran to face trial.

Though there are other places, like Lebanon, Palestine, Venezuela which will also have the right to put some of the same people on trial..Those rogue states really do attack a lot of other countries.

Comment Re:Maybe I'm missing something (Score 1) 150

There's a very big difference between chess, and things like programming. In chess, you can quickly and easily tell whether you have checkmated your opponent or been checkmated. It's possibly to algorithmically work out what is correct, with absolute precision in the simpler cases. With programming, there is no state of correctness. LLMs are based on the idea that you must want some kind of weighted average of the sum total of human endeavour they have ingested. That whole technique can only get you so far, even when you get it 100% perfect.

Comment Re:The real death of the internet (Score 2) 44

It happens because big corporations are above the law. Remotely activating filters on things people have bought fairly in a shop unless they show you some ID is obviously a crime in violation of the Computer Misuse Act. But Apple are a big corporation, so nobody will be prosecuted. Like nobody was prosecuted when Sony included trojan horse software on music CDs, back when music CDs were a thing. Like nobody is prosecuted when Microsoft says WIndows 10 will be the final version of Windows, people buy computers based on this assumption, and now they're being made obsolete if you want to connect them to the internet. It's all because we live in a plutocratic system.

Comment Re:Code review is not what AI is being sold as (Score 4, Insightful) 41

In general, the principle problem with LLMs is that they're completely unreliable, due to the basic design. But in cases where they''re just saying "look at this, maybe this is a problem", reliability is not required because if it makes no sense, someone can just say "no". The problem comes when people begin to trust them, despite them being completely untrustworthy. Applications where trust is not required are fine.

Comment I don't get how this kind of thing works (Score 1) 122

I'm too wily to buy something like that. But if I did buy one of those things, and an advertisement appeared on it, my reaction would be "how dare you put advertisements on my fridge which I paid money for! I'm not buying anything from your company now unless there's absolutely no alternative!"

Advertising-supported free services? Fair enough, I understand the bargain there. Showing advertisements on something that's paid for? You are now my enemy.

Comment Re:NIST algorithms (Score 1) 68

If you see no threat from quantum computers, then remaining completely classical for simplicity and performance is indeed the consequent choice. But whether there's a threat from them is a very hard question to evaluate since predicting technological development is very difficult, so one shouldn't jump to suspicion simply from there being people making the opposite judgement.

Comment Re:NIST algorithms (Score 1) 68

They're fine against conventional attacks, I don't know where you're getting that idea from. It's true that one finalist turned out to have a catastrophic flaw, that doesn't mean they're generally worthless. What it actually means is that we have less confidence in their primitives than we do in conventional primitives, since they haven't been around for nearly as long. Whether you think the increased risk of a break being found is more or less than the risk of someone attacking your conventional algorithms with a quantum computer is a judgement call.

However, you don't have to make this judgement call. You can use hybrid algorithms - sign/encrypt with both conventional and quantum-resistant algorithms. Then in order to break your crypto, an adversary needs both a good quantum computer and a PQ algorithm break.Or for signing, you can use SLH-DSA, for which the only primitive is well-tested hashes, providing its slowness and large signatures don't make it impractical for you.

Comment Re:Interesting but not exciting (Score 2) 52

To get the intended effect, you need to go into Mission to the Unknown expecting the Doctor to turn up at some point, watch The Myth Makers having no idea what that was about, and finish up watching The Massacre wondering if the whole thing is coming to an end. Over the course of 20 weeks, naturally. It's a period when the show deliberately abandons having the individually named episodes neatly arranged into different stories.

Comment Re:The evil BBC deliberately DESTROYED Dr Who (Score 2) 52

Doctor Who in particular, has not only generally been progressive, but sometimes quite radical. Take one of my favourite 60s stories for example, The War Games, which starts by looking like it's set in WWI before revealing a larger sci-fi setup, turns the commanders of both sides of a number of wars into aliens with more in common with each other than the soldiers they command, directing the war by playing board games with each other, to which the solution is for all soldiers everywhere to unite with each other in collective mutiny against their commanders. Rather than having a military command structure with captain as protagonist, it has a hero who hates guns and typically defeats enemies representing power by being more clever than them and refusing to take them seriously.

Slashdot Top Deals

Kleeneness is next to Godelness.

Working...