If you've never worked through a derivation/explanation of the Y combinator, definitely find one (there are many across the internet) and work through it until the light bulb goes off. It's pretty incredible, it almost seems like "matter ex nihilo" which shouldn't work, and yet does.
It's one of those facts that tends to blow minds when it's first encountered, I can see why one would name a company after it.
$GOOG is 2 or 3 times what it was before the AI boom, depending on when exactly you define "pre-AI boom", so this isn't quite the full story. I tended to think Google was undervalued in the early 2020s and people weren't giving enough credit to how dominant e.g. YouTube was, so maybe it's accurate now and Google won't have as strong an AI correction even if one happens.
It’s sitting at ~29 forward/trailing p/e which means that it’s likely to drop 30% if there’s a correction and even more if there’s a broader economic thing going on that causes ad spend to go down.
Something under-appreciated: If you pretend a company is paying out 100% of profits as dividends (which it theoretically potentially could, and is useful as a financial modelling tool), then the inverse of P/E, E/P, is an interest rate on the price of the stock.
Ideal P/Es thus shouldn't be flat, they should be tracking long-term bond rates. This isn't an empirical observation, just a theoretical one of what "ideal" should be. But one should rationally expect P/Es to go up when interest rates drop.
It is disappointing to me that even Shiller doesn't really consider this much.
People have said this, but so far if anything the opposite has been empirically true. OpenAI had a huge lead and it just didn't matter, Anthropic and Google both caught them and now they're neck and neck. It seems like compute overhang forecloses the possibility of runaway progress which eliminates all your competitors.
Any feedback process has a hard threshold for instability. The PA system doesn't howl until the microphone is close enough to the loudspeaker. The atomic bomb doesn't explode until the fissile material reaches critical mass. If you don't know where the threshold is you can't extrapolate.
Compute is a limiting factor now, but there have already been huge improvements in compute efficiency, e.g. mixture of experts. It seems extraordinarily unlikely that there are no more to be found. And compute capacity continues to increase too.
Multiple package managers are trying to move to ssh keys and other stronger forms of verification, as well as trying to outlaw binary tarballs and other such things. It's somewhat slow going: package owners sometimes get grouchy about this and drag their feet.
Which is wild we’re coming full circle. Everyone made these things easy to publish to so we could onboard newbies faster but then we all figured out that sacrificing security to save someone 10 minutes of reading was a bad idea.
Don’t get me started on everyone being git@github.com…
Not quite the history as I remember it. These package managers were often created by small teams of people who originally didn't know they'd turn into Microsoft acquired corporations. The intent wasn't to onboard newbies. People just didn't have a reason to use insanely targeted attacks on OSS Things because OSS being used in such a widespread manner wasn't as common as it is today. It really feels like people have forgotten how things were back then.
> Multiple package managers are trying to move to ssh keys and other stronger forms of verification, as well as trying to outlaw binary tarballs and other such things.
What does binary tarballs have to do with it? Whether the package is in source form or in binary form, a long dependency chain attack is equally likely to succeed.
Altman keeps on telling people he’s going to take away their jobs. He says that because it gets cred in tech circles, but in America this is an existential threat, not much different from telling someone “I’m going to break your kneecaps”. Of course some subset of people are going to respond with violence.
The sheer tone-deafness of AI marketing is going to come back to bite us very hard. This is probably just the beginning.
Yeah part of me thinks the reason we know all their claims are bullshit is because you’d have to be pretty dense to think that you could promise eliminate >50% of jobs in many high value sectors within 12-18 months and _not_ expect to create more than a few people who’d have nothing to lose…
Hasn’t Creative Commons disavowed or at least really downplayed the NC license for exactly these reasons? There are so many ambiguities and headaches involved that the only advice I’ve ever seen is not to use it.
- If it’s a company doing an NC license, probably best to be careful because they can make your life hell with lawyers.
- If it’s a random joe doing an NC license, feel free to ignore it because they don’t have the money to defend it anyway. Especially so if it’s CC-BY-NC-ND, people that pick that one are especially likely to be in the all-bark-no-bite category.
At least that’s how one of the companies I worked for treated CC licenses… I don’t work there anymore.
Likewise, I was unaware of this (and still see it in use regularly, especially on places like Printables as I've recently gotten my hands on a printer myself)
A lot of those bugs were found by seasoned developers and security professionals though. Anthropic claims that Mythos is finding vulns from people who have no security background, who just typed "hey, go find a vulnerability in X", went home for the night, and came back the next morning with a PoC ready. They could definitely be an exaggerating, but if it's true that's a very different threat category which is worth paying attention to.
Previous models have done this just fine. For the last year, whenever a new model has come out I just point it at some of my repos and say something like "scan this entire codebase, look for bugs, overengineering, security flaws etc" and they always find a few useful things. Obviously each new model does this better than the last, though.
Imo that's a big deal primarily because the issue with automatically discerned vulnerabilities has long been a high volume of reports and a very bad signal-to-noise ratio. When an LLM is capable of developing PoC exploits, that means you finally have a tool that enables meaningfully triaging reports like this.
I agree that I miss when spaces could be on a grid in Snow Leopard instead of only in a straight line, but what is wrong with Exposé? From my POV it works the same as it always has.
IIRC Expose is now called Mission Control (four finger swipe up on my system).
My problem with it is that it's useless if you got more than few windows open - the preview is just too small to actually see which window you are after (it's all padded for the looks). IMO if they actually used tiles, potentially grouped by app - it would be so much useful.
Yabai looks cool tho, but requires so much permissions and potentially disabling system integrity protection that IDK if it's a go for me.
I don't have a Snow Leopard Mac in front of me to compare, but as far as I can tell, four finger swipe works the same way it always has. It's hard to find targets if you have dozens of windows open, yes, but I don't remember old Exposé being any better in that regard. I think one or two macOS releases had some kind of "Mission Control UI" in the bottom half of the screen when you did the swipe, but that's gone now and the entire screen is used for the tiles.
Can you explain more about what regressed since the old Exposé? I'm just not seeing it.
Yes, the stupidity and shortsightedness of American Catholic integralists like Vermuele is stunning to me. If America does ever become a Christian theocracy, it's going to be a Protestant theocracy. It wouldn't be an altar-and-throne continental monarchy, it would be more like Cromwell's England, where "Papists" were considered enemies of the state. Do these guys not remember that Jack Chick wrote just as many comics villainizing Catholics as he did atheists? That's how evangelicals actually think, once any temporary alliances of convenience have accomplished their goals.
It’s going to be Evangelical. Some variety of megachurch prosperity teaching that faults the poor with some kind of republican ideology.
That’s why anyone that believes in separation of religion and state should tell these folks anytime they push for Christianity in schools, just tell them: ok but it needs to be the true Christianity- Jehovah Witnesses- then they will shut up. They hate Jehovah witnesses, then Mormons, then Catholics, …
It stuns me that Republican Mormons think that Evangelicals like them for anything but their political assistance. As soon as Evangelicals remove the non-Christians, their tent will get smaller, just like you're saying.
I have Mormon family that thinks that they're welcome in the Evangelical tent (they'll even visit the Ark Experience!), but Evangelicals hate Mormons just like they hate gays, liberals, trans people, atheists, etc. It's just that Mormons (for now) vote the way that Evangelicals want.
At this point, it's going to be some sort of post-Christian, culturally Christian social media influencer-driven, conspiracy theory-laden melange that incorporates everything from Tartarian giants to simulation hypothesis to Flat Earth. Q Gospel indeed.
There are so, so many naked scams right now which are still being hyped by Big Names, it boggles my mind. Substrate is an obvious scam, data centers in space are an obvious scam, it goes on and on. But they all have famous names insisting with a straight face that these should be taken seriously. It feels like psychological warfare against intelligent people.
Quantum is a scam. It's embarrassing to see it being peddled as the next big thing, with public companies being valued in the billions. For technology that does not exist in any useful sense.
It's one of those facts that tends to blow minds when it's first encountered, I can see why one would name a company after it.
reply