his talent was the ability to identify a demand and invest in it. he didn't even "build" the thing, he bought it ready made and simply changed the terms of service so that the demand could be satisfied.

millions of people spending time and money on that platform could surely invest their talent and money on something more constructive too, but that's their choice.

'are you not entertained'?

de-anonymizing will just mean that bots will use false/fake identities. one would hope that another far more effective approach would be considered: educate people in critical thinking, rendering bots, fake news and propaganda merely a nuisance. my hope is slim though not because it would also be a (very) tough endeavour but because a population able to think critically and resilient to misdirection is the last thing that powers that be want. can't have it both ways, and de-anonymization definitely isn't about bots or the children, it's about controlling the discourse and the herd.

valid concern, but there's a reason it only works on the pixel. it indeed may be backdoored but the same is true about any other device, or even more so, and security is essentially a cat&mouse game and a series of tradeoffs.

about funding google ... also a valid concern but overstated. still if you have to favor any company at all then google in all its evil is the one which has contributed more by a very large stretch. promoting a handful more pixel sales in a niche market has a far smaller impact than e.g. promoting apple by making it linux compatible, or promoting windows with wine. the tradeoff is that these initiatives promote alternatives aswell.

you really seem more obsessed and desperate as usual, which was already quite a lot. take care, good luck.

the end user delivers encrypted data and decrypts the result. if you trust the encryption method then you can trust the output.

now, i have no idea how this system in particular works and what its capabilities or intended use are (didn't rtfa), but for specific operations and given specific assumptions this is entirely possible.

just "disable javascript on this site".

it wasn't a new contract but an expansion of the first, but that doesn't really matter: if anthropic refuses to remove these restrictions then the rational outcome is that they would stick to the first contract with the restrictions in place (unless pentagon wants to void it, which they could simply do) and walk away from the expansion (or second contract). similarly, the pentagon might simply disqualify anthropic for expansion or further contracts, and is free to void the first one under the conditions it itself agreed to. i don't see a problem here and nothing here implicates a "supply chain risk" in any way, it's a bogus accusation.

the pentagon might very well think: i don't trust these people for whatever reason. well then go find another supplier, right? and optionally argue what exactly the problem is supposed to be?

the core of the "supply chain risk" argument is the idea that for some reason anthropic must honor any demands the pentagon feels like making at any point in the future even after an agreement has been made, if anthropic refused, that would be a "risk". well, tough luck, but that's just not how contracts work. it was a mere bullish threat to force them to comply, anthropic didn't yield and walked away and now trumpteam follows up with the threat just the same, maybe out of spite, maybe to save face ... or maybe because the intent was destroying anthropic all along.

it seems they did:

source: https://www.ox.ac.uk/news/2026....

for the emperor(s)!!

businesses should be doing that anyway, openjs is just providing a new service to assist in migration. to me it sounds more like trying to appear relevant. businesses that do sensible maintenance will not need such a service, businesses that don't probably won't bother with this either. i'd say a less glamurous but more productive effort would be funding audits and backports of security patches to older versions.

updating is a risk in itself and aged software isn't necessarily more risky as long as it is reasonably maintained. outliers get a lot of attention but the biggest exposure and highest risk is new releases and then current versions. do update, but do it with care, and as late as reasonably possible. also, stay away from products or components without a good track record of maintenance.

it can't drag on for 4+ years because the world's economy would get back to 1800 levels. well, china and russia would likely weather that out.

it will probably end when trump manages to weasel out proclaiming he obliterated something and won, but the iranians don't seem to be inclined to give him an offramp so easily for now. they should ask for total sanctions lifting, dismantling of all us bases in the region (that's symbolic, it means cleaning the rubble that's left from them in any case) and the recognition of a palestinian state. plus reparations.

now, if it drags on very long iran's current political system could potentially collapse sooner than that. i don't think the iranian ever nation will, and any case the us regime would probably collapse sooner than that. fafo.

the legal framework for war is quite clear. respecting and enforcing it is another matter. the mechanisms to get authorization are often subverted (which means those actions my be technically legal but still fraudulent) and even more often are ignored altogether.

the us war against iraq in 1991 had a mandate from the un security council, which is the sole international authority to authorize military action, thus was a legal action according international law. the us tried to reuse this same mandate to invade iraq in 2003 but the un explicitly ruled this illegal. the un charter was signed in 1945 and has authorized armed intervention in about 7 other instances (many on dubious grounds, but still legal) so except for those all other us wars from there on were straight out illegal according to international law.

according to us law, the constitution gives the sole authority to declare war to the us congress. the last time congress did this was in 1942 in ww2. in 2001 congress gave a statutory authorization of military action against any country which "planned, authorized, commited or aided" the 9/11 attacks. this authorization was used for aggression on several countries but all evidence of "planning, authorizing, committing or aiding" by any of them was fabricated and phony (if provided at all), so those wars weren't really covered by that statutory authorization. meaning (unless i'm missing something) all us wars since 1942 with the exception of iraq 1991 were also illegal under us law, except nobody ever bothered to figure that out.

both parties agreeing to anything has little bearing on legality. the un charter does ofc allow military action in self defense, but that's a gray area and there is rarely consent about who started hostilities. the right to preemptive strikes has been invoked quite often too but to my knowledge in no instance the legal requirements were actually met (not even remotely because they're very strict), as of today the un hasn't sanctioned a single instance of preemptive action in the world.

so it's basically a "might makes right" scenario with occasional nice words or lies thrown into it. as clausewitz put it, war is a continuation of politics by other means.

strong the cognitive dissonance is!

know what ... i think i'll keep expressing my opinions freely on the open internet, thank you. if that bothers you, you're welcome to ignore it ... or to vomit your bombastic deathwishes to clearly show where you morally stand, that's fine! note i'll also keep using the word "antisemitism" in its universally accepted meaning according to which absolutely nothing i said can even be remotely related to it and cheerfully ignore this pathetic attempt at subverting language in order to slander and censor. have a nice day!

apparently they used palantir maven to recollect intel data with which then claude produced target lists with coordinates, ordinance recommendations and legal justifications. this process would usually take weeks but allowed to list and strike over 1000 targets recently in iran in a single day. supposedly the targets are verified by a human official, although verifying 1000 targets with any rigour in 24h seems a bit much.

incidentally, it has now transpired that the strike on a school that killed over 100 girls aged around 12 was most likely by the us rather than israel. maybe the officer needed a coffee break. since claude is actually more accurate than gpt i guess we should expect more of this. on the flip side llms seem to be easy to deceive, they also blew up quite a few helicopters painted on tarmac. otoh, the war is illegal (it's getting even embarrassing to use that word) and murderous to begin with and striking civilians has been israel's deliberate tactic of choice for decades. so this is just cost efficiency. the tools change, the barbarism stays the same, death multiplies.

the restriction is that depending on what you use you will be sued into oblivion anyway. copyright is about big money, just don't mess with big money.

imo ai is just a tool. going to this extreme is just as absurd as prohibiting photographs to be copyrighted, because the images were actually captured by a chip or a photosensitive emulsion and not a pencil or a brush. well, there's also the use of the tool and the intent. someone decided the framing, the motive, the mood, whatever, to express something. similarly, with ai someone crafted the prompt with something in mind, that's the creative process. the tool is revolutionary, but the concept is essentially the same, this is just about big business protecting their rents.

in our chaotic and psychotic current world i'm not very surprised that the chinese of all people are the only ones who understood this correctly: in china you can copyright ai-generated content as long as you can prove creative process or effort. which is really the sensible thing to protect (if you still believe copyright is there to protect creativity, which is ofc a myth from disneyland (pretty literally)).

well, you can also copyright ai-generated work in the uk, but not for long because that's just an oversight and they're working on turning that around, so hurry up!