I think this is a good outcome. We need to maintain a lowish-cost manufacturing partner and India is better than China.

I don't agree. Charging an up-front price for a good product is not what enshittification is. That is simple price competition. Gradually reducing the value of the product by introducing ever more-annoying little profit streams is what enshittification is.

This is for now a competitive market, and these quotas a pretty generous for what one license is marketed to be used for.

The "summary" (which I guess is just the whole article these days) says "TSMC will make AI5, which just finished design, initially in Taiwan and then Arizona."

Offhand it's hard to think of a captcha can distinguish between an AI and a person using common peripherals on untrusted hardware. And it won't get easier.

It would be funny if we had a law for an AI safeword - the AI must respond to a given phrase with a given response. This should be doable for the mainstream cloud-based AI that most people use, not because an AI can be shown to do anything reliably by itself, but because a wrapper could be put around it to ensure that's what it does.

Of course, all bets are off for self-hosted AI, or whoever has no reason to comply with US government rules.

Why is it "another" reason? Is Samsung on a naughty-list I don't know about?

Regarding "dumb money" I have wondered where the smart money is going now that tens of millions of people have millions of dollars in index funds for their 401k, intentionally using no information to make the decision to put more in all the time, and striving *not* to be responsive to the underlying economy at any time - just buy and hold.

What it makes me wonder is where the smart money is going. Let's say somebody is so "smart" they can consistently beat the market (perhaps because they have a stream of insider information from a few difference sources). Do they play the normal market but with outsized performance? Do they instead form a private equity firm? Or...?

Intuitively it seems like the very wealthy might have a new private financial playground now that the open market so democratic.

The problem is that Linux isn't 'superior.' Does it do some things better? Sure. Maybe a lot of things.

The vast majority of computer users simply *don't care.* Nor do they need to.

Windows works well enough, it does anything almost anybody wants 'a computer' to do, and you don't need to put too much thought into it.

Well, no, but on a per-dollar basis - a billion is a very, very big denominator. Or in other words their per-hour income is hard to justify on the basis of anything like 'merit.'

It's really hard to say how much a winner has 'produced' compared to the alternative future in which they didn't exist. Bezos worked long and hard, without profits for many years, to bring Amazon about. He succeeded and a lot of us choose to buy a lot of stuff from Amazon all the time. What if he had not. None of the other online retailers would have grown to fill that niche? Perhaps not to the same degree, since arguably that is why Amazon beat them. But to what degree would that hypothetical competitor have fallen short? Certainly their value would be a large fraction of what Amazon's now is. Yet our system rewards the winner as if all that wealth simply wouldn't have been created... not proportional to merit or effort, but "winner takes all."

I do believe this. You don't just gradually and smoothly age all the time. Besides cell-level degradation (which I now learn is not constant either) you have a health setback, like an illness / chronic condition, or significant injury, and then never fully recover to whatever level of "youth" you had before.

Be interesting to see this play out in a more western and liberalized culture.

I live in a US state that is poor but gets most of what it does from oil money. I applaud the transition to renewables but it don't see how the state government will get by.

They should just call their experiment a side effect of trying to money, and then keep going until/unless it is banned, while they bribe congress not to ban it. That's how it works, is it not?

From the right: "Biden did this!" sticker on every gas pump calling attention to higher gas prices

From the left: "Study finds tonnage estimates overstated in greenwashing conspiracy!!"

It is just impossible to make a CO2 regime that won't immediately be pilloried from the right and left.

Well yes that would be easy to do. If somebody uses an LLM to play chess directly, they're doing it to push the envelope of the 'LLM' or let's call it "an AI using LLM techniques with a generalized mechanism" that might work in a more basic modality to provide language generation and more. But telling it the rules of chess should still be done using language I would hope.

I agree it ought to integrate with an OS undelete. Some filemanager-type programs have their own 'recycle bin' (recoverable trashcan) which does work but then you have a whole new mechanism to worry about.

It works on bottled CO2. That's a limited natural resource. What if we run out?