HNHacker News
TopNewBestAskShowJobs

impossiblefork

2,372 karma · joined January 6, 2022

submissionscomments
impossiblefork··on Meta’s new commercial terms will allow the scraping of news sites: experts
Maybe, maybe not.

But in navigating to the page being requested the user may have agreed to something, and that may be a contract.

impossiblefork··on Meta’s new commercial terms will allow the scraping of news sites: experts
Yeah, but you can have one page where people agree to a certain license and if they agree are allowed to download another document.

TOS agreements often have terms concerning things like in which venues disputes about them are to be handled etc., and often have penalties other than the service simply being ended, and if those are valid, then this kind of thing is too.

impossiblefork··on Pornography showing strangulation and incest to be banned in UK from November
The strangulation thing very clearly is harmful, so I think that part is good, maybe the incest stuff is, maybe it isn't.

I hope the ban is well-considered, but a ban on strangulation and choking in pornography is a good idea. Allowing production or sale of media reinforcing something that dangerous would not have been super smart.

impossiblefork··on Meta’s new commercial terms will allow the scraping of news sites: experts
If you don't have a license to download the material, aren't you doing piracy?

If the licenses say that download is only permitted if you agree not to train on it or similar, then if you have downloaded it, you have agreed to the license, have you not?

impossiblefork··on Banks are now trying to offload the AI data center debt they just underwrote
I think the data centres stay, continue to be used for LLMs but end up not getting the intended returns and possibly ending up in the hands of creditors who don't end up with the sort of money they expected.

I think at least we Europeans will really want these data centres once prices become reasonable though. We really need training infrastructure.

impossiblefork··on Switzerland proposes banning profiling-based ads to minors
Oh.
impossiblefork··on Switzerland proposes banning profiling-based ads to minors
Ah. Here in Sweden basically all TV is now cable-based. We apparently still have some wireless stuff, but I haven't seen it in years, so I've basically forgotten how TV used to work, and still works in some parts of the country.
impossiblefork··on Amsterdam police confiscates server left-wing ISP A/I at American request
Light source, not lens.
impossiblefork··on Switzerland proposes banning profiling-based ads to minors
But people pay for TV? Surely, if ads weren't there, the TV production would be funded by the people paying.
impossiblefork··on Switzerland proposes banning profiling-based ads to minors
Yeah, I can understand an exception for milk producers.
impossiblefork··on Nobel Peace Prize for 2026 to Navanethem Pillay
I actually don't think it is.

I think Timur, Genghis Khan, etc. and a bunch of people of that sort are over the line.

impossiblefork··on Switzerland proposes banning profiling-based ads to minors
Here in Sweden it's illegal to make ads directed at children. Consequently the TV channels that do so transmit from abroad.

This would of course have been easy to ban anyway-- it's not like we can't have rules on foreign internet firms, for example, but historically the politicians were unwilling to make a useful law.

I think the Swiss could take one step further and add a general ban on ads targeting young children in addition to this; and then we Swedes should imitate the package.

impossiblefork··on Strengthening our defenses against distillation attacks
Also, Mistral Large 4 will probably beat GLM 5.3 once post-training is done.

Relatively similar concept, but has more total parameters, more active parameters and ordinary full attention instead of deepseek sparse attention, so it should probably be GLM 5.3-beating.

impossiblefork··on Just say no to OpenAI and Anthropic
If there were an International AI Club which trained models and all the members sat down each year and voted for the board I would trust it even if it had a cutthroat commercial arm.

I think the only thing that matters is control.

impossiblefork··on Nobel Peace Prize for 2026 to Navanethem Pillay
It's happened before though, that seemingly insignificant things motivate people to do this (https://en.wikipedia.org/wiki/Hitler%27s_reference_to_the_Ar...), and specifically with the Armenians again, which is, I guess, a sort of irony:

>Our strength lies in our quickness and in our brutality; Genghis Khan has sent millions of women and children into death knowingly and with a light heart. History sees in him only the great founder of States. As to what a weak Western European civilisation asserts about me, that is of no account. I have given the command and I shall shoot everyone who utters one word of criticism, for the goal to be obtained in the war is not that of reaching certain lines but of physically demolishing the opponent. And so for the present only in the East I have put my death-head formations' in place with the command relentlessly and without compassion to send into death many women and children of Polish origin and language. Only thus we can gain the living space (lebensraum) that we need. Who after all is today speaking about the destruction of the Armenians?

The speaker of the above quoted words is Adolph Hitler.

impossiblefork··on Nobel Peace Prize for 2026 to Navanethem Pillay
Well, if I didn't think it was interesting I wouldn't post it. Then of course you reassess and start to wonder whether that initial remark really was worth it, and then your comments begins to evolve, and then you have done this for years and it becomes a style where the evolution of the comment expresses uncertainty and causation.

I always avoid genuinely changing the comments though, appending only, and ceasing to append in case someone responds before I'm done.

impossiblefork··on US proposes $100k charge for international students to do post-graduate work
They would actually. But it wouldn't be the "EU government phone". It'd be Jolla, and then competitor after competitor would form.
impossiblefork··on Nobel Peace Prize for 2026 to Navanethem Pillay
Yeah, I can be slow in getting them finished.

But at that point I can't have mentioned the ICJ stuff, can I? Hmm. Oh well. I get what I get for my editing.

impossiblefork··on Nobel Peace Prize for 2026 to Navanethem "NAVI" Pillay
How do you mean, and who?

If you mean Swedish state television, what I meant was that any such major judgement by a UN court like the ICJ, would be front page news.

Remember, this stuff was subject to long and incredibly annoying thing in the Security Council as well, and SVT ignored that too. It was the biggest thing that happened that whole year, basically and the entire coverage by SVT were those words I quoted in the comment before this, that "Armenian separatists have agreed to leave Azerbaijan".

impossiblefork··on Nobel Peace Prize for 2026 to Navanethem Pillay
You remember Machado though...

Peace prize for someone who was literally arguing for subjecting her own country to an invasion and foreign occupation.

I remember how Swedish state television suddenly flipped very after that, and started cuddling Cuban exiles who wanted the same there. I think this whole thing was certainly coordinated somehow, behind the scenes in order to attempt to legitmize a future US invasion of Venezuela and their general conduct in the Carribean and South America /Pacific.

Things like this happened during expulsion of the Karabach-Armenians too. Then SVT went out and said that "Armenian separatists [had] agreed to leave Azerbaijan", and they left out the artillery bombardement, the deaths by starvation and the ICJ judgement against Azerbaijan's closure of the Lachine corridor, even though such a judgement by an organization like the ICJ would otherwise by first-page news.

So while I can't be sure that there's some NATO-internal press office coordinating how planned future military activity by members is to be presented in the media, it seems that Turkey could achieve that, and that the US could achieve that.

impossiblefork··on OpenAI cannot make AI safe on its own [pdf]
Surely the fact that the foundation part of OpenAI has at least some power should make them able to feel that the firm is supposed to at least care.

I like to think that, because OpenAI have pledged to ensure that AI benefits all of humanity, and has also aligned with the US DoD, which is obviously killing at least some people, who thus will not benefit, is effectively asserting that all people who are not US persons are not part of humanity and I think this is what they actually believe (and I think this latent belief together with their tendency towards EA/the corporate version of LessWrong will mean that they'll try to murder us EU AI researchers once we start making breakthroughs-- it's simply an unavoidable consequence of the ideology-- to them we're simply Werner Heisenberg, and their writings and fiction and their view of the stakes make that clear), but even if this harsh assessment were accepted they would still be able to expect the board to care about safety.

impossiblefork··on Mistral Large 4
It obviously increases the time by 20 times given fixed hardware, it probably also makes the hyperparameter search more difficult and expensive, depending on how predictable the performance of your architecture is.

This was pretrained for 2 months, so a model with 1T active parameters would take 3.33 years. Maybe a 4 months pretraining run can be acceptable in a future where Euclyd or VSORA machines have been rolled out in quantity. But we don't know how far off that is.

But that (i.e. models made with 4 month pretraining runs) would probably allow an EU-made genuinely China-beating open model, so if VSORA is rolled out in quantity, let's say, tomorrow, then this could be a thing that would happen this very spring.

impossiblefork··on Mistral Large 4
We need to maintain the capacity to build reasonable models in the meantime, until we have hardware though.

Remember also that this model is still in preview and won't finish post-training until in a month or so.

impossiblefork··on Mistral Large 4
A 1T model with quite few active parameters though.

A 1T model with more active parameters is obviously more expensive. If it had 1T active parameters it would probably be 20 times more expensive to train.

impossiblefork··on Sharing AI progress in mathematics
Yeah, that's one of the big things of TCS. I think I see that as bigger than that Millenium Prize problem.
impossiblefork··on ICC judge on what U.S. sanctions mean for her and global courts
Sorry, but understand -> understanding and "he isn't sitting" "without its own powers" instead of "he's sitting" and "with its own powers".
impossiblefork··on ICC judge on what U.S. sanctions mean for her and global courts
Having seen some other texts, at least one other judge (Guillou) has really strange attitudes of the role of the court, discussing the European project, etc. in articles about the court.

The fellow does not give the impression of understand fully that he's sitting on an internal court with its own powers, and the way US/Polish/etc. crimes have been excluded from the Afghanistan investigation even though people relevant in such an investigation are particularly easy to interview since so many are in Europe I think these people genuinely do not understand their job, or have given themselves another job in their minds.

Prost is now basically doing public outreach, Guillou was doing public outreach and diplomacy trying to interact with the EU to get it to enforce its blocking statute regarding sanctions on the court-- all this stuff instead of using the powers they're granted by Rome statute to deal with this specific problem.

They have the remedy. Anyone retaliating against the court can be ordered arrested, and the signatories have to do it or they're in breach of the treaty.

The level of subservience of the judges is just ridiculous.

impossiblefork··on GPT-Synopsys: Frontier Intelligence to Revolutionize Chip Design
Eng: So I prompted Claude to turn the 125 microservices into a single Erlang program. Better not tell the client.
impossiblefork··on The AI Race Just Got Awkward
Yeah, and Anthropic probably got inspired to this new fast read-in thing for making agentic stuff make more sense from the latest DeepSeek model. Maybe it was in the pipeline, but it clearly has the same effect and DeepSeek had published it by the point Anthropic dropped their prices for reading tokens in, so they may well have copied it.
impossiblefork··on US sanctions force The Netherlands off Microsoft and toward alternative NixOS
Consider though system like the Swedish system which existed in the Viking age with Things and a king.

Things were courts with huge power, and consisted of a gathering of all free men and had basically unlimited power, because they were the people. Meanwhile, the government was the person of the king.

I think that's an example of a system where the courts (Thing) is clearly not part of the government.

← PreviousPage 2 of 34Next →