not2b

You changed the article's title to an incorrect one. The tree of primitive Pythagorean triples is ternary, not binary. Each node has three children.

My guess is that it would be much more effective to capture and remove CO2 directly at the source, for example at a cement plant. While this could be done at a fossil fuel plant as well, it seems a lot less attractive: you give back most of the energy you just got from burning the fuel.

Note, though, that "the government" (NIST to be specific) says that requiring passwords to be changed every 90 days is counterproductive and shouldn't be done, yet many corporations (including my employer) still mandate it. Corporate bureaucracy can be as backward and counterproductive as government bureaucracy.

The following argument from the blog post is wrong:

"First, write a bash script P1 that runs a SAT solver on every possible SAT problem and halts if any of them take exponential time. You don't actually have to run P1, just call HALTS(P1, Solver) to see if the solver solves all problems in polynomial time."

The difficulty here is that it could turn out that there's a SAT algorithm that is better than exponential, but the SAT solver used in the experiment does have worst case exponential time. What has to be proved, to prove P != NP, is that there is no possible SAT algorithm that has sub-exponential worst case time.

No. Canada passed a law requiring Facebook to pay news media for links. Meta said no, we aren't going to do that and banned news instead.

You can argue that this was a predictable response by Meta or that it was a stupid law, but it was not a ban.

That can be a threat, but a billionaire American or South African with similar power and motivation is also a threat.

Instead of the laser focus on TikTok as a threat, it would be better for the US and Canada to have real data protection laws that would apply equally to TikTok, Meta, Google, Apple, and X. What the EU has done is far from perfect but it bans the worst practices. The Chinese can buy all of the information they want on Americans and Canadians from ad brokers, who will happily sell them everything they need to track individuals' locations.

Perhaps the way to get anti-regulation politicians on board with this is for someone to do what was done to Robert Bork and legally disclose lots of personal info on members of Congress/Parliament, obtained from data brokers and de-anonymized.

Verilog has a sliding-window part select, which produces a slice that has a fixed width. It's an lvalue expression (it can be assigned to). It's ternary.

array[addr +: W]

or

array[addr -: W]

The former chooses W elements beginning at addr, the latter chooses W elements ending at addr.

That discussion and similar discussions that seems to come up regularly was probably the reason they don't want HN traffic, and it's unfortunate the the moderators here haven't worked harder to shut that stuff down. The gender of their developers just isn't relevant to the quality of their work.

In many cases they are doing their own designs (done by engineers they hired away from the various electronics companies) and having either TSMC or Intel fab the chips for them. They can design specifically for their workloads to get better price/performance than they can buy from AMD or Nvidia.

AMD is way behind. But one kind of competition that's going to more important in a year or two is that the huge companies (Meta, Google, Amazon) are working on their own AI chips, designed to run their workloads at lower cost (time and power) than they'd get from Nvidia. So data centers might wind up full of non-Nvidia chips, and Amazon may make them available to their customers. But they will probably be less flexible.

Here's what Meta says they are doing: https://ai.meta.com/blog/next-generation-meta-training-infer...

A parallel that was not mentioned has to do with intellectual property. In the 19th century the US did not respect foreign copyrights or patents; smuggled British machinery was cloned to produce the American industrial revolution, and Charles Dickens was the most popular author in the US but he didn't get a dime from American publishers, who could just take and print his works.

Likewise the Chinese often ignore foreign copyrights and patents, though not as much as the US did back then.

For the first example, for C or C++ code gcc will catch the dangling reference/pointer and warn about it, and since my group normally turns on -Wall -Werror, it's an error. However, the analysis is local and not as powerful as Rust's borrow checker.

Try int* bug() { int longlived = 12; int* plonglived = &longlived; { int shortlived = 13; plonglived = &shortlived; } return plonglived; }

With gcc -Wall -Werror

The most blatant whoppers that Google's AI preview makes seem to stem from mistaking satirical sites for sites that are attempting to state facts. Possibly an LLM could be trained to distinguish sites that intend to be satirical or propagandistic from news sites that intend to report accurately based on the structure of the language. After all, satirical sites are usually written in a way that most people grasp that it is satire, and good detectives can often spot "tells" that someone is lying. But the structure of the language is all that the LLM has. It has no oracle to tell it what is true and what is false. But at least this kind of approach might make LLM-enhanced search engines less embarrassing.

If a takedown notice is sent, maybe it is because someone's rights are being infringed, or maybe someone is just trying to censor something they don't like, or falsely claim authorship. You seem to be assuming that the fact that the notice was sent means that the sender is definitely a victim and the owner of the repository is definitely the perpetrator. Not always.

The notice and takedown part of DMCA was a reasonable attempt to address a real problem. The main flaw seems to be that bogus takedown notices are never prosecuted, so there's a lot of abuse.

Messier catalogued many objects, some of which were galaxies, but he did not know what they were. It was Hubble who first figured out (in the 1920s) that there were definitely separate galaxies.

https://lco.global/spacebook/galaxies/history-discovery/

The reason that it is impossible is that there is no possible set of answers that would require the test-giver to acknowledge that a test-taker passed the test. Anyone the test-taker does not like can be failed.

I notice that the terms "debt" and "bankruptcy" (and their variants) appear nowhere in the article.

It may be that one LLM had text from the language in its training set along with translations, and the other didn't.

Microsoft forced a lot of uniformity on x86 designs, with standard methods of probing to find all of the buses and devices, while historically ARM devices where embedded and the architecture was arbitrary, which was why you needed device trees. The kernel couldn't figure out the device configuration on its own.

It won't happen because neither the US nor the EU antitrust regulators will allow it. It isn't just up to Qualcomm and Intel.

Some other deal, where Qualcomm buys some pieces of Intel, might be possible.

I should have qualified this with "the engineering departments at UC Berkeley". Everything we put out (papers, technical reports, open source software) was on the Internet. Formats were varied; LaTeX and Postscript were commonly used. PDF a bit later.

Everything was digital at UC Berkeley back in the early 1990s and before.

Even before generative AI was an issue, in the US you couldn't just tell a court "this is a photo, therefore it is evidence". There had to be witnesses would could testify about how that photo was produced, and the other side could cross-examine.

But yes, you're already seeing politicians caught doing or saying embarrassing things claiming it's all a deepfake, but in most prominent cases of this there were lots of witnesses who can confirm that yes, they did say that.

The number is being reported accurately, but with only one digit of significance. Rounding 39 to 40 isn't lying and isn't deceitful.

They might just change its status to a vanity top level domain like ".lol" or ".sucks" and sell it to the highest bidder, taking the money. They would justify it by saying that they want to promote stability, maybe require that the new owner honor the domains at least until they expire and then charge what they want. That seems to be the way ICANN works these days.

One of the massive platforms, like DraftKings. They won't even notice. The point of all this is the effect of making sports betting legal.

Pretend to be 1,000 people, and each sock puppet bets $10k.

If OnlyFans is taking a 20% cut and Apple is taking a 30% cut of transactions on their platform, it seems that Apple is 50% more toxic than OnlyFans.

I've never done business with them and am not interested in buying that kind of content, but it certainly seems like an improvement over any more traditional sex-related work for those who are interested in being in that market.

If OnlyFans is really paying 80% of gross to creators, it seems that any kind of "regular professional status" would be worse for them, at least for the top 10%.