The minimal size limits of reasoning abilities are not clear at all. It could be that you don't need all that many parameters. In which case the door is open for small focused models to converge to parity with larger models in reasoning ability.

If that happens we may end up with people using small local models most of the time, and only calling out to large models when they actually need the extra knowledge.

It does seem like a cynical attempt to make more money.

Even if someone were to game resolution to make a market resolve the wrong way, it wouldn't affect how the market priced the probabilities in the time before the resolution.

Of course you have to read the fine print very carefully to understand what a particular market is about.

The more deceptive thing is the potential disparity between the title and the rules. Most market participants read the rules closely (although some do not), and so the title + headline probabilities may be not as they seem.

If Linux were to contain 3rd party copyrighted code the legal entity at risk of being sued would be... Linux users, which given how widely deployed Linux is is basically everyone on Earth, and all large companies.

Linux development is funded by large companies with big legal departments. It's safe to say that nobody is going to be picking this legal fight any time soon.

For example let's say you want to make an app that every day writes a backup to a particular location e.g. 1Password can do a daily backup of your encrypted passwords to a backup location.

Or, let's say you want to make a GUI around a command line program that stores its config as a dotfile.

Without a way to save access to file system locations persistently, apps would be forced to constantly shove open panels in your face all the time.

It's not the bait on HN that you need to be worried about but the propaganda from your own government.

Apparently nobody at Microsoft considered that blocking critical software hurts Microsoft more than the open source developers being blocked.

You can definitely use git as a backend for building such a system, but some extra tooling is necessary.

The article is written as if it's by some crazed conspiracy theorist. It's dripping with confirmation bias at every turn. Tiny coincidences are seized upon to confirm the author's suspicions whereas other explanations are minimized.

This kind of one-track reasoning almost always turns out to be wrong. Did the author accidentally stumble upon the truth? It's possible. Back has all the right qualifications to be Satoshi. But so do many other people.

> In keeping with this belief, Mr. Back made his Hashcash spam-throttling software open source.

> Satoshi did a similar thing. He released the Bitcoin software under M.I.T.’s open-source license, which allowed anyone to use, modify and distribute it without restrictions.

The numerous observations such as this only seem impressive to people who don't know anything at all about the subject. Occam's Razor suggests that the reason that such irrelevant observations were included is because Carreyrou doesn't know anything at all about the subject.

> When we compared those errors with the writings of our hundreds of suspects, Mr. Back was a clear outlier. He shared 67 of Satoshi’s exact hyphenation errors. The person with the second-most matches had 38.

The article does not improve.

Open source developers are doing Microsoft a big favor when they support Windows and publish Windows builds and installers. It's a substantial effort, and apparently that effort isn't appreciated.

If all open source software dropped support for Windows, it wouldn't really affect the open source community that much. It would definitely cause headaches for Microsoft however.

That's not to say the rescue mission was wrong, the psychological advantage of Iran capturing the pilot would have been immense. But, it demonstrates just how weird the US military is.

Most militaries would have had no choice but to let the pilot get captured and then negotiate a prisoner swap at a later date. The US had the option to mount a rescue mission, and merely having that option available is a strategic disadvantage. Now Iran knows that the US is very unwilling to suffer captures. Now Iran is incentivised to maximize captures in the case of a ground invasion.

The US could probably win with a ground invasion, if they committed all their forces. But they're definitely not willing to suffer the consequences, so the effect is that they cannot win[0]. The US Army is a supremely powerful force that nevertheless cannot be used offensively anymore because the US is unwilling to suffer the consequences of doing so, kind of like a nuclear weapon.

[0] it reminds of Feynman's anecdote about a stage hypnotist. When the hypnotist invites you on stage and tells you that your eyelids are heavy and you cannot open them, you are aware that you could open your eyes if you wanted to. But in front of the watching crowd, you of course "choose" to obey the hypnotist and keep your eyes closed. So were you really able to open your eyes? The US military "chooses" not to open its eyes.

GPT-5 is good at benchmarks, but benchmarks are more forgiving of a misaligned model. Many real world tasks often don't require strong reasoning abilities or high intelligence, so much as the ability to understand what the task is with a minimal prompt.

Not every shop assistant needs a physics degree, and not every physics professor is necessarily qualified to be a shop assistant. A person, or LLM, can be very smart while at the same time very bad at understanding people.

For example, if GPT-5 takes my code and rearranges something for no reason, that's not going to affect its benchmarks because the code will still produce the same answers. But now I have to spend more time reviewing its output to make sure it hasn't done that. The more time I have to spend post-processing its output, the lower its capabilities are since the measurement of capability on real world tasks is often the amount of time saved.