somenameforme
And while I do expect progress to eventually slow, there's no reason to yet think we're at the plateau point or necessarily even especially close, so they're probably only going to get better. But at this point, even if they didn't I'd still be saying this. 'Programming' with an LLM is a loathsome and unpleasant task, but that's because it's no longer programming but instead an entirely different design-centric domain where it's handy to have some programming knowledge to nudge it in the right direction sometimes.
I think human programmers are going to go the way of human computers. [1] And I'm obligated to link to that because now most people probably no longer even know such an occupation existed, even though it rather obviously must have if you think about the past for a moment.
If a big pharma had the choice between an expensive treatment that could help treat any cancer, or an extremely cheap treatment that could prevent any cancer - which would they pick? The answer when you're motivated by profit is unfortunately different than the one when you're motivated by social benefit. And this is more than just a thought experiment because this issue is going to drive which type of treatments are pursued.
This is often especially cynical because [popular thing x] is often something that neither party wants to pass. So a very recent example of this was an act against insider trading by politicians, that was bundled with a voter id law. Neither side wants to clamp down on insider trading from politicians, so the entire act was DOA. It was nothing but bread and circuses.
By contrast the LLM 'skill' of 'sit on ass and talk' is something basically everybody already does. So you just need some minimum level of general cognitive ability and you're good to go.
Like is there going to be anybody, whatsoever, who's genuinely aiming to go long on Anthropic? It just feels like this massive collusion that 'everybody' is in on where we all know it's going to crater but the whole game is to pump it as much as possible and just try not to be the last guy holding the bag.
Obviously LLMs are going to revolutionize the world, and they arguably already have, but there are open weight models available for free with frontier level performance, and the price to run these models keeps plummeting which, in turn, is already driving token costs down. Like many things now a days, this is all highly entertaining, yet absolutely insane.
What I was responding to is somebody above thinking that Microsoft was being "generous" with this deal - what I was saying is that they were almost certainly just sticking to greediest possible behavior that stays within the bounds of the contract. You can describe billion dollar multinational corps in quite a lot of ways, but generous isn't one of them.
I'm not privy to a standard Microsoft contract so I can't say for 100% certain, but I'd be quite shocked if they don't provide a clause that workers can only be terminated for cause. So they're legally required to buy out contracts if they want to get rid of a substantial number of workers in a short period of time.
Churchill: Madam, would you sleep with me for $1 million?
Woman: Why Winston, yes I would.
Churchill: How about $10?
Woman: What sort of woman do you think I am?
Churchill: We've already established that. Now we're just negotiating the price.
Something I'd emphasize is that the main reason that China/India are rising is simply because relative tech disparities between nations are fading rapidly. And once you have at least remotely comparable tech, the size of your economy is going to be heavily dependent upon your population. The US has 340 million people. China and India have about 1400 million. That's a losing battle. But notably the EU has 450 million. That's well below China/India, but well ahead of the US. That should count for something, but for some reason it seemingly does not.
The reason these states returned them in exchange for a few bucks and empty assurances (which is an tautology - the difference between 'assurance' and 'guarantee' is precisely that there's no legal obligations or bindings with assurances) is because the alternative was being deemed a potentially rogue nuclear state which would have seen them rapidly invaded, very likely be a coalition of nations - viably including the US as well, and disarmed before they could manage to reverse engineer the protocols.
And yeah, Ukraine would not have been invaded if it had its own nukes, neither would Iran. It's also highly unlikely that the revolution, coup, or whatever you want to call it in 2014 would have happened if Ukraine had nukes. That involved things like John McCain giving speeches encouraging Ukrainians to overthrow their Russian-leaning president right in the center of Kyiv. [1] That was a massive power play that probably would never have happened with a nuclear state. Similar for the assassination of Soleimani in Iran, and so on. In any case, if a country wants to guarantee its sovereignty, it needs nukes.
And the reason this guarantees sovereignty is because every single country is aware of Proud Prophet [2] even if most people are not. It was a large scale nuclear wargame simulating all sorts of scenarios for 'peace through force' with nuclear weapons. In every single scenario it resulted in the defacto extinction of the Northern hemisphere. It led the US to turn away from escalation and hostility and towards reconciliation and negotiation with the USSR. Within a decade the first McDonalds would open up in the USSR, and shortly thereafter the entire USSR would collapse with Donald Duck being held up as a sign of freedom.
Software is a very clear example on this front. Valheim [1] is a game that was made mostly by one guy, at first at least. Not that long ago, this would have been a multi million dollar game produced by a wide array of people. It's just that tools, markets, and more have made it possible for even single developers to have an absurd degree of efficiency. And LLMs are probably going to bump this up exponentially.
This is why a metaphor I'm currently drawn to is one of people living in a very agrarian society just prior to the Industrial Revolution. What would they have envisioned being created? Super farms of course. And those super farms were indeed created, but the overall impact was such that those super farms became a practically inconsequential part of the overall economy. The future we're moving into isn't going to be the same but more efficient for those on top. That efficiency is going to revolutionize the economy in ways that are probably impossible to see from a distant, much how somebody living in an agrarian society couldn't have even begun to imagine the 'real' impact of the Industrial Revolution.
If you have a high enough degree of certainty somebody could have a 99% chance of winning with an expected 50.1%/49.9% split. For the split to mean anything you need to know the confidence/variance.
As for Europe, this [1] is a graph of the EU's share of the global GDP, PPP adjusted. It's been in steady decline for decades. And so discussing the decline of Europe, at least relative to the rest of the world, is hardly outrageous.
[1] - https://www.statista.com/statistics/253512/share-of-the-eu-i...
It also creates a scenario where people remain on the bench long after their competence has likely faded. RBG, for instance, was 87 years old when she 'retired.' Her final wish is to not be replaced until a new President had been sworn in, which was predictably not granted. She had 30 years to choose that presidency herself, but simply refused to do so.
I'd also observe, vice versa, that after Nintendo jumped on the bandwagon of charging players to use their own internet connection that their sales started declining significantly. I don't know if it's causal; even people who chose to opt out themselves might not even really know. What I mean is that this feeling of getting nickeled and dimed is going to imperil fun in some impossible to quantify fashion, and if these various actions hit another completely unquantifiable breaking point, then you start losing players in yet another unquantifiable way - until suddenly you notice that people just don't seem to want what you're offering anymore.
--
EDIT: And I had a chuckle seeing a topic just posted: "Disney Raising Prices Days After Confirming Ads Will Now Show Up on Every Tier" [1]
'Shareholder value' is another problem. Shareholders, in general, do not hold shares of companies because they want those companies to succeed, but because they expect returns from those companies. If they can get a 50% return and then have the company die in a fiery blaze, most shareholders would be perfectly fine with that so long as they were reasonably confident that they wouldn't, themselves, get burnt. So you have shareholders with no interest in the long-term similarly aligned with executives who have no interest in the long-run.
There's also probably more technical issues like discounting earnings over time. A penny earned today is generally considered significantly more valuable than a penny earned tomorrow by companies due to inflation, opportunity cost, and other such issues. So even if you think the long-term is relevant, the expected earnings cost would have to be dramatic to make it comparable to the impact of next quarter.
But from the PS4 onward, the consoles were made of a relatively lightly modified retail hardware that was sold for profit as quickly as possible. Both the PS4 and PS5 were being sold for profit (on the hardware), less than a year after launch. I think this was due to again companies squeezing a shrinking playerbase even harder, to try to maintain profit levels.
I think the obsession on next quarter, while pretending that there's no such thing as long-term consequences, is eating companies alive. Because there are absolutely long-term consequences.
---
PS1 = 102 million units sold
PS2 = 160
PS3 = 87 million
PS4 = 117 million
PS5 = 95 million (June 2026)
---
These numbers would be bad alone, but two things make them especially awful. The first is that there's no meaningful competition anymore. XBox has completely dropped the ball whereas the PS2 initially had the original XBox (which wasn't great, but had novelty going for it), Dreamcast, GameCube, and others all viably competing in the same approximate domain.
The second is that when the PS2 launched in 2000 the US population was 281 million. Now it's 349 million. I'm using US as a proxy for rest of the world, instead of the global population, since we're mostly interested in the developed world as that's the market for game consoles. So the break-even point in terms of marketshare relative to the PS2 would be 349/281*160 = 199 million units sold, and that's if we assume significant competition.
So not only are people leaving consoles, they're leaving in droves. And that leaves the companies squeezing their shrinking player bases even harder, which will make people leave even more quickly. They've lost around 50% of their adjust playerbase over the past 20 years, and at this rate I doubt game consoles will even exist in 20 more years.
This is also why I think the 'acceleration' hype is probably unjustified. What seems vastly more likely is that we start to reach an asymptotic 'perfection' where models can accurately and consistently provide meaningful next token prediction with minimal hallucinations. And then a race downward to compact this into the most efficient package possible.
This would mean the models still would not be able to create things that do not follow in a reasonable probabilistic fashion given a knowledge of [everything], and so that is the domain, 'applied creativity', where I'd expect humans to remain dominant.
And at that time it was Ukraine rejecting any calls for a ceasefire without major concessions from Russia. They wanted the immediate and unconditional return of all land as a precondition to even beginning to negotiate, where they would have sought further concessions like reparations to agree to peace. And beyond that they were making statements, in public - let alone private, such as "[Russia] must be defeated, be subjected to a painful defeat, as painful as possible." [1] These are not the words or actions of a country that'd just agree to peace, without major concessions, if they were trending towards a potential absolute victory. But of course no country would. It's the nature of war. The side with an edge will expect significant concessions to end a conflict.
And, obviously a general mobilization would cause major backlash within Russia, but they'd almost certainly be just fine. There's also no such thing as a "verified death count." Again going back to our propaganda list: "7. We suffer few losses, the enemy's losses are considerable." On top of both sides actively lying, they also likely don't even really know themselves, especially Ukraine. The overwhelming majority of their military is made up of forcibly conscripted people, many of whom would flee at the first opportunity. They certainly know the dogtagged KIA, but likely simply have no idea exactly how many of the missing are fertilizer, and how many managed to make their way to freedom. They also have a motivation to not know, to defer the KIA compensation payments.
[1] - https://www.euronews.com/2022/05/22/us-ukraine-crisis-podoly...
Nonetheless, this doesn't mean they'll be identical anymore than you'd be identical to yourself if you started in the world today once again. And of course also our children aren't just us, but a mixture of both their parents, with a bit of randomness tossed in for fun.
And of course the reality is that your children are a mixture of two people, with a bit of randomness tossed in for fun.