While I wouldn’t put it past Google/YouTube to do something as shitty as this, I think people are far too quick to assume foul play over the much more likely possibility that the world’s largest video platform occasionally shits itself.
The issue is that profit-driven hospitals will look at a model that can greatly assist in diagnosing certain things and go “Cool, we can fire those specialists and have some pre-med run the model and make an accurate diagnosis”.
I nearly had a heart attack until I read the word “ARM” at the end.
It seems like ARM Arch is doomed for whatever reason.
My country, which has killed more people than any other geopolitical entity in human history and is ruled by a literal monarchy, is clearly free and egalitarian, unlike that horrific autocratic nightmare of…
checks notes
…a tiny Central American nation that’s been routinely bombed and exploited for its entire existence.
I’m begging liberals to remember the one cool thing they ever did, which was chopping off their monarchs’ heads, and commit to that. Just that. Come on now, this is an argument that’s been solved for like 200 years now: monarchs and aristocrats are dogshit and deserve to be forcibly removed from their positions of wealth and authority.
and packaging this into a system that meets the scale and reliability requirements to make it commercially viable hasn’t been reproduced to date
Your overall point about EUV being difficult isn’t wrong, but this line is really where the typical liberal forecasting of China’s capabilities fall apart: they don’t give a shit about it being commercially viable, they give a shit about having the industrial capacity.
The reason why EUV is more or less a cartel monopoly in the West is that it’s a cobbled together collection of scientific principles that work well enough that the first few companies that figured it out could make insane profits off of it, and then proceeded to patent the shit out of it to prevent anyone else from doing so. The engineering behind EUV is… not great from a reliability standpoint, most notably the fact that EUV has an average downtime of something like 10% (meaning your fabs are offline 10% of the year for maintenance), in large part because you’re shooting little droplets of liquid metals with a high intensity laser which tends to splatter and require cleanup. There are potential alternatives to this process for creating the kind of UV light you need for lithography, such as particle accelerators, that are theoretically superior but the R&D into those alternatives costs tens of billions of dollars with no guarantees that any of it will ever become profitable, so Western capital doesn’t bother trying.
China doesn’t have that profit restriction. It needs the ability to produce bleeding edge chips to remove its reliance on an increasingly hostile West, and it has not only the engineering and scientific power to brute force that kind of R&D but the ability to devote a sizeable portion of its national resources to doing so. It doesn’t matter if its profitable, it matters if they’re able to decouple a critical industry from the West and ignore sanctions accordingly, and that has infinitely more value than a shareholder dividend, so they will put the resources into doing so and, inevitably, they will figure it out. And from what we’ve seen over the past 2 years since the trade wars have started, they’re not only succeeding but doing so ahead of expectations, in large part because increasing tensions have made life a living hell for Chinese scientists and engineers abroad working in these industries due to racism and suspicions of spying which push them to emigrate back to China and lend their expertise there instead.
In 20 years, chips made in mainland China will be competitive or even superior to their Western counterparts unless the West undoes 50 years of neoliberal rot overnight and replicates what the CPC is doing for silicon manufacturing or the CPC collapses and China experiences the same shock doctrine that the former Soviet states did in the 90s, and neither of those outcomes look likely right now.
There’s a top surgery joke in here somewhere, I can feel it.
Sure there is. An enormous chunk of housing sits unused and empty because real estate speculators want to rent them out at exorbitant prices rather than use it for it’s intended purpose of having a roof over people’s heads.
Pass nationwide legislation that restricts owning housing for commercial purposes beyond a certain threshold, and put rent controls in place pegged to 20% of the median income per town/city. You’d eliminate 95% of homelessness before the ink was dry, massively increase homeownership rates, and be the most popular politician of an era.
It’s not even an ebil communist plot, and it’d still be more effective than giving even more money to private developers on a pinky promise they’ll build something people can afford, just trust them this time.
Steve Jobs quite openly hated the idea of anyone gaming on a Mac because he felt like it made their products seem more childish or something. It seems like either nobody at Apple has managed to dig that particular brainworm out yet or have just decided that printing iPhone money makes all other concerns irrelevant.
You have to sell a new phone every 12-18 months, because otherwise the shareholders eat you alive for not chasing infinite profits. You have to differentiate your new phone from your last phone, even if there are no meaningful changes to be made and the last phone was good enough for everything anyone would ever use it for (as was the one before it, and the one before that, and etc etc). You have to push for people to buy the new phone, because otherwise you don’t make money.
So you tell the engineers to bump up the clock speeds on the processor 5-10% so you can market it as being faster. You market the phone as being revolutionary for using the USB connector that was forced on you by regulators because your proprietary one was filling landfills with e-waste and pretend like it was your brilliant idea all along. You make sure to limit that USB connector to speeds that were outdated 10 years ago purely so you have a built-in ‘upgrade’ for your next phone where you fix the thing that shouldn’t have been a problem to begin with.
And then you realize your phone overheats because you overclocked the processor, all to squeeze extra performance out of a chip that 99.9999999999% of users will never notice or need. You’ve made the user experience of your phone worse purely so you could pursue an untenable goal of endless profit, a pattern you will repeat every 12-18 months for the rest of eternity or until the climate wars claim your life.
Only the most sane and functional economic system.
We can only do terrible things permanently. Good things are always temporary and we are powerless to change that. Children have to starve because otherwise politicians might have to actually do some work for a change. This is a very good and serious political system and we should just accept it as it is and ignore any alternative.
That’s what 70 years of exclusively using your military as the enforcers of neo-colonialism does. Turns out what works for leveling a low-tech guerilla hideout in Vietnam or Afghanistan isn’t so effective when your opponent has comprehensive SAM networks.
Well, I’d rather the day be sooner than later.
Agreed, but we’re not the ones making the decision. And the people who are have two options: move forward with a risky, expensive, and potentially career-ending move with no benefits other than the system being a little more maintainable, or continuing on with business-as-usual and earning massive sums of money they can use to buy a bigger yacht next year. It’s a pretty obvious decision, and the consequences will probably fall on whoever takes over after they move on or retire, so who cares about the long term consequences?
You run months and months of simulated transactions on the new code until you get an adequate amount of bugs fixed.
The stakes in financial services is so much higher than typical software. If some API has 0.01% downtime or errors, nobody gives a shit. If your bank drops 1 out of every 1000 transactions, people lose their life savings. Even the most stringent of testing and staging environments don’t guarantee the level of accuracy required without truly monstrous sums of money being thrown at it, which leads us back to my point above about risk vs yachts.
There will come a time when these old COBOL machines will just straight die, and they can’t be assed to keep making new hardware for them.
Contrary to popular belief, most mainframes are pretty new machines. IBM is basically afloat purely because giant banks and government institutions would rather just shell out a few hundred thousand every few years for a new, better Z-frame than going through the nightmare that is a migration.
If you’re starting to think “wow, this system is doomed to collapse under its own weight and the people in charge are actively incentivized to not do anything about it,” then you’re paying attention and should probably start extending that thought process to everything else around you on a daily basis.
Translating it isn’t the difficult part. It’s convincing a board room full of billionaires that they should flip the switch and risk having their entire system go down for a day because somebody missed a bug in the code and then having to explain to some combination of very angry other billionaires and very angry financial regulators why they broke the economy for the day.
Company whose business model is based entirely on running an enormously massive and expensive LLM and then serving content with it publicly for free with no greater ideas on actually turning that into a business is going under. In other news, water still wet.
I’ll admit I thought the AI bubble was going to last longer than a few months (and inevitably FAANG will probably artificially extend it until even they have to admit there’s not a ton of productive real world uses for it), but I suppose late stage capitalism has to speedrun the boom-bust cycle as it gets increasingly desperate for profit.
There’s the other half of this problem, which is that the kind of code that LLMs are relatively good at pumping out with some degree of correctness are almost always the bits of code that aren’t difficult to begin with. A sorting algorithm on command is nice, but if you’re working on any kind of novel implementation then the hard bits are the business logic which in all likelihood has never been written before and is either sensitive information or just convoluted enough to make turning into a prompt difficult. You still have to have coders who understand architecture and converting requirements into raw logic to do that even with the LLMs.
Continued use of the swastika in his military and a stubborn refusal to acknowledge how many Soviets the Finns starved to death in Leningrad while working with the Third Reich, presumably.