yyyk

"we used a fuzzer to minimize a corpus of 10 million PDF files down to 4,200 without any loss of code coverage"

Did they need a fuzzer for that? They could've render them all and see what's exercised?

Trump publicly prohibited various Israeli operations (at end of 2025 op and 2026 Iran war), publicly badmouthed Nethanyahu repeatedly (via Barak Ravid), and had various diplomatic initiatives Nethanyahu didn't like. It's pretty obvious who has more influence here.

WSL1 is generally better for these usecases.

gci /w than?

If they're good enough at useful tasks and cheaper than any human, the economic and political effect is the same regardless of whether they are 'intelligent' or 'pretend-to-be-intelligent'. The philosophical debate does not change the practical effect.

in the name of personal power and wealth.

Technological innovation is perhaps enhanced by capitalism**, but is not dependent on it or a result of it. Development would have happened anyway given current technological levels, just in a different form, and the race between states would have led to deployment, even if possibly slower deployment.

** There's an argument that Google hindered AI deployment for awhile because the CEO was worried about its effects.

In my now long lifetime, I can't tell you how many times I've heard the phrase...

Russell's Turkey Parable:

"The man who has fed the chicken every day throughout its life at last wrings its neck instead, showing that more refined views as to the uniformity of nature would have been useful to the chicken."

My money is my own affair. Now talk to the point instead of the person.

We can all read the news and see AI constantly improving at a very quick pace. What's the barrier AI can't improve beyond? One that would last a lifetime? By all means (since you're such an accomplished person), explain yourself.

The problem is that he relies on a dubious Acemoglu estimate, without realizing that at best it's temporary. AI will be better than humans in doing tasks (the qualifiers don't matter in the aggregate). Any jobs then would be bullshit jobs, and everyone will know it.

There are several good points there, but there are also several points where I must disagree.

First, Acemoglu may have a Noble Prize (occasionally dubious data selection aside[0]), but even he does not pretend to have the information of a central planner nor does he think that central planning is good idea. If actual businesses, which have local information, believe AI helps them, I'll bet the actual businesses are much more likely to be right, no matter what calculation Acemoglu did or how many Nobels he has.

Second, it is likely that AI will eventually be better than (nearly all) humans in most economically useful things. We can debate the timeframe but it will happen and likely not that far away. That Federal Job Guarantee would generate bullshit jobs, and he won't be able to hide that from people. Ultimately, we'll reach an economy which objectively does not need humans and everyone will know it. He needs to face that reality and overcome it, and not hide behind temporary and dubious estimates.

Third (admittedly a minor point), while the criticism against EA etc. is very justified, it's not quite fair to blame them (overwhelmingly STEM people) for not reading where the humanities did many of the same errors earlier (the author points out some of these) and discredited themselves. The people who could have taught them to not do it failed to teach themselves.

And fourth: I'm pretty sure a lot of company would be able to charge AI agents themselves. The new economy will not be dead, it just won't involve (many?) humans.

[0] https://xcancel.com/joefrancis505/status/2059340591490552054...

That's easily worked around (e.g. fisheye and appropriate correction). The real problem is that the scenarios discussed don't allow for setup, otherwise they could just use preplaced cameras...

While the argument is probably right, the author may notice he's asking AI websites on AI water usage. Perhaps he should consult more neutral sources, at least to introduce the habit of critical thinking.

It looks like that by simply reducing use in welding, lifting, and purging gas (all with clear alternatives) and maybe also 'leak detection' and 'other' (not expounded on in the article), they can fill in for the entire Qatari output, and that's without including extra production and recycling which is quiet possible.

There are ways to shield data centers if one is serious about it...

e.g.

https://www.reuters.com/world/middle-east/oracle-opens-first...

But without a nuclear weapons program, the entire sanctions regime wouldn't have started (yea, I know today half of those are anti-terrorist sanctions, but that's not how it started, it was morphed later on). It should be considered as part of the losses.

Iran claims 1T USD damages as a result of US leaving JCPOA alone - in 2021. Now add in 5 more years, wars, sanctions before JCPOA was signed, direct expenditures on enrichment...

https://www.aljazeera.com/news/2021/2/21/us-sanctions-inflic...

A treaty whose key articles would start expiring in.. late 2025. Which Iran had no motivation whatsoever to extend had it being kept (imagine this Iran but with 2-4 trillion dollars more, more than a few going to drones and missiles). You'd have this war but on way worse terms.

Yea, the US joined in in 2025, what should it imply about a future war? The assumption that Iran doesn't know who's bombing it sounds rather dubious. If anything, it should be very much in their interest to assume away US involvement unless 100% proven, given fighting an additional enemy tend to be very bad and US is so powerful. Unless...

Maybe the strategic balance creates a situation where it's advantageous for Iran to pull US in regardless of non-involvement. They don't do well against Israel alone (see rather low damage of 4 separate large scale attempts at attacking Israel directly), but US is so much easier to pressure via the Gulf. Indeed, this scenario doesn't quite need Israel.

So US risked getting pulled in not due to attacking in June 2025, but because the cheque given to the Gulf was starting to expire, the power balance was objectively swinging in favor of Iran at the location where Devereaux sees as the most important part of the Middle East. Now, say there are powerful states who feel they are in a decent position now but also that the strategic balance would slip away. What do they tend to do? Devereaux can consult his WW1 history.

We saw regime collapses in the Arab Spring - it's not a simple or short process, most regimes survived (either directly or via reversion). Even when a regime was overthrown, the replacement was usually not more hostile to Israel. e.g. Syria isn't more hostile than it was. Thing is there isn't all that much 'fury' since Arabs already assume the worst of Israel, while reasons for relative peace remain as is or are actually strengthened by the revolution process (e.g. economy, desire for quiet following violent revolution, new regime wanting to establish itself, etc.)

Look, it's a done deal. Some of the choices Wayland made are not to my liking, there will be a long term cost (even static linking won't save you from differing protocol implementations). But it's done and there's no point in complaining.

(Running X11 right now, I'll switch when the distro forces me to, in hope I'll get a bug free experience after everyone else runs it)

Everyone Iranian leader who is killed is post-facto named 'moderate', with very little proof of being moderate while alive.

Nobody forced Iran to bomb gas and oil fields that aren't even of their enemies, and to start doing it before Pars was bombed (e.g. Shah gas field was targeted 2 days ago). Place the blame on Iran who is actually 'wrecking Asia' and not the ones fighting it.

I think the idea behind a prediction market is pretty interesting...

But we're in an era of less and less responsible government oversight

The era doesn't matter. The incentives are so bad, eventually you'd get a horrible result even with oversight. These markets are prone to this and shouldn't exist.

continuous compression

'Solid compression' (as WinRAR calls it) is still optional with RAR. I recall the default is 'off'. At the time, that mode was still pretty good compared to bzip2.

Even then, there were better options in the Windows world (RAR/ACE/etc.). Also, bzip2 was considered slow even when it was new.

Except here the entire Iranian strategy (what's left of it) _relies_ on targeting civilian 3rd parties. Without threatening Gulf oil, they've got nothing.

"fighting back" == blowing 3rd party civilian installations.

First, JCPOA restrictions would have started to end this year, the 2031 situation was just extra bad.

Second, had NPT been enough, why was JCPOA necessary? Because NPT was not near enough to stop a determined state. NPT 'limits' do not limit enrichment. They just state enrichment must be done for civilian purposes and leave no enforcement mechanism. JCPOA also had no real enforcement mechanism starting in 2026. Iran could have simply taken all the money, buy weapons (legally), reach the legal max (which is basically infinity in 2031, a bit less earlier), and pressed forwards immediately. There was simply no way to enforce (some people think _this_ operation is too risky! Now lets give Iran over a trillion dollars and years to prepare).

Third, you're just wrong about current negotiations. Iran explicitly demanded 20%[0], beyond the JCPOA and with no civilian use.

Look, JCPOA debates always end in the same way. The advocates are asked how the original deal was enforceable in any way past 2031 (or even 2026) and they have to deploy word salad because the actual answer is 'Iran could have easily built a nuke, inspectors just allow us to see it but they would have been no way to do something about it'. It would be more honest if many of them just admitted they wanted to allow Iranian nukes.

[0] https://xcancel.com/laurnorman/status/2028050672583618946

Doesn't harm the mouse. But would it harm the normal human cells?