ajross

Late to the game, but I have to call out your shameless cherry picking. That elipsis in your wikipedia quote skips: "which triggered a stock market crash and bank runs in several countries." I mean, dude...

Yes, obviously, the subprime crisis was long in the making and has deep root causes. But the bank runs of 2008 and the frozen global financial system were 100% preventable, as they were a confidence shock (not a fundamentals thing) stemming from the sudden uncertainty due to Lehman.

A little assistance from the Bush administration in 2008 would have very plausibly produced a much softer landing and required a lot fewer bailouts overall. This really isn't controversial.

"Let them fail" is almost always the worst thing to do.

> Who controls this controls me and that's not OK

I'm lost, are you arguing against Binance or against freezable crypto? Both "control" you! There are like three billion dollars out there in presumably-north-Korean-hands that their owners thought they "controlled". They didn't.

So given the choice, yeah, I'd like some safety please.

That's the logic that led to "Let's just let Lehman fail" and precipitated the 2008 crash. At scale, markets are not zero sum and failures sink all boats.

There are no good answers here. But pointing fingers and assigning blame is for sure one of the worst ways to do the analysis.

I'm just saying that the scale of those cuts matters in the analysis. How many billion dollars of paychecks have been frozen?

> Freezing crypto is so disgusting to me I don’t know…

The feature literally appears to have just saved over two billion USD from being stolen! I mean, fine, "disgusting" is subjective. But the need for the feature is real and quantifiable (and very large) and it did what it advertised.

> LuaJIT handles double, strings, and dictionaries at near C level performance.

Yeah, but again you don't reach for C when you have a string or hash-based logic problem, which is the reason that other platforms are favored. You reach for Python[1] to solve those problems, and it works fine.

You reach for C when you care about cache footprint or memory layout. You reach for C when you need SIMD parallelism. You reach for C when you care about doing the math in single precision or 32 bit integers and will handle the precision worries yourself.

You emphatically don't reach for python (or Lua) in those situations. But notably you can reach for numba, and implement your small core of performance code in python next to the other logic you already have. And there's value there.

Screaming about using Lua instead is just nonresponsive. Lua is arguably better at performance code than Python. Lua is absolutely not faster than numba, nor will it ever be (absent a numba-like translation layer) and repeatedly trying to argue so is very much missing the point.

[1] Or fine, Lua or JS or whatever. But I'm trying deeply not to let this turn into a frothing language flame, and this is after all a python discussion.

Well, we all (hopefully) agree this should be legally protected.

But no, that's silly. The folks running these protests are the same population that's been protesting law enforcement in other contexts forever, and ACAB is a founding principle. They know as a matter of unshakable faith that the fuzz will cheat and lie and commit atrocities, and the the paradigm of the protest is to catch them in the act.

So failing to apply that to your IT work is... bad praxis, sorry. The one argument I can see here is if they actually wanted to get caught to troll an article just like this one. Which would make the operator in question a martyr and not a dummy, I guess.

These benchmarks and rankings seem more noise than signal these days. They're uniformly divorced from sincerely-attested real world experience. And the real world experience isn't driving any consensus either.

So here's a meta-benchmark: at what point does the act of choosing an LLM product for a reasoning task become less reliable (in the sense of "predictive of future success") than hiring a human being for that same task? Have we already cross that threshold? I think we might have.

And once we do... what happens then? Will the truly elite models be hidden in a sea of mediocrities simply because the goofball "managers" doing the hiring are too dumb to recognize their genius?

> When OP wrote "inline Cython" they didn't seem focused on numerics, they talked about "proper C speed"

Isn't that basically what I said? Lua isn't "proper C speed", nor is PyPy, the cancelled Python JIT in question, nor even V8. That's a different problem area with different technologies. The fact that you like coding backend stuff in Lua or whatever isn't really saying much about whether numba is great.

But it is. Numba is great.

Yeah, this is sort of the disconnect. Lua is a fast dynamic interpreter, sitting well above Python but rather below e.g. V8 in its ability to run scalar code over "interpreter types" like doubles and strings and vectors and hash tables.

The upthread comment was about an idea for "inline Cython", which was then hijacked because that's largely what numba is. These are technologies aimed at optimized native code running over native types to do stuff like numerics and ML and other HPC tasks. Lua isn't any more responsive to this domain than Javascript has been, it's not really the subject under discussion.

That's not actually true, though. In some sense all governments are posessed of "weapons" like this, the cost of use being only the violation of some norm. Or even law. The "rule of law" is a gentelemen's agreement at its core. Yet the world is filled with basically good governments who broadly follow that code!

So the way you treat that is by calling out and resisting the attempts to weaponize the government in new ways, not just throwing up your hands because it was "eventually" going to be used anyway. No, it's bad. You should say that it's bad and take action to oppose it.

Not really the bent of the article, but it's also worth pointing out that this shows some really terrible opsec practice by the activists in question. Like, no, you don't just spin up a cloud server to coordinate your protest activity and then ssh into it from your home IP. Dude...

There's no "selling" of data. These are just connection logs kept as part of routine auditing/anti-spam/anti-abuse practice by the ISP (something we all agree here is a good thing). And ICE appears to have gotten a warrant for them, which showed them the personal IPs of their targets.

You can argue the warrant was poorly justified (and I'd agree!), but I don't see a plausible world where "ISPs aren't allowed to do connection analysis on their own networks" is going to help anyone. Well, anyone except spammers. They'd love that.

Both-sides-ism is probably the single worst stance to take here. You're tacitly siding with whoever actually holds and wields the power. Not caring which one wins isn't helping anyone at all, even you.

> Saying it can't be fast ever is a huge cop-out.

Maybe, but the world is still waiting for the counterexample. Python is simply not a runtime well-suited to optimization.

In the meantime, those of us who actually enjoy real HPC problems are using numba and rolling our eyes about all the confuffery around JITs. Maybe one will arrive that proves us wrong, but numba is fast today.