The problem is that hyper-advanced automation is essentially unpriceable.
I mean, I could point you to a few economics journalists who would argue otherwise. Ed Zitron’s been screaming about the downfall of AI for the last two years straight.
But the market can remain irrational longer than you can remain solvent, as the saying goes.
It’s an entire industry shilling a fantasy on the back of some very impressive sales demos.
Doesn’t hurt to have a bunch of tech-friendly goobers running state and federal governments who are turning out the taxpayer’s wallet to finance these hallucinations.
I mean, I could point you to a few economics journalists who would argue otherwise. Ed Zitron’s been screaming about the downfall of AI for the last two years straight.
I feel like you’re only reading every other sentence of what I say. In this instance, you seem to have fixated on this part, but sailed right past the part where I said that there’s zero evidence that anyone can actually produce hyper-advanced automation. I never argued that it was a rational decision to go all in on this possibility, and that’s entirely clear from my previous comments.
Ed Zitron is completely correct, but he’s also making exactly the same argument I am; that these people cannot actually achieve the technological revolution they are promising. That doesn’t change the fact that, if their wish granting genie was real, it would basically have unlimited upside. The problem is not how they’re pricing the outcome, the problem is how they’re evaluating the probability of achieving the outcome.
there’s zero evidence that anyone can actually produce hyper-advanced automation
There’s plenty of evidence (Waymo, for instance, or Xaiome) that companies can produce “good enough” advanced automation. Tesla’s just not the guy to make it happen. Same with AI. Very possible to produce useful tools (Alibaba’s Deepseek, Insilico Medicine’s Pharma.AI, etc) with LLMs. Sam Altman’s just not doing it.
Ed Zitron is completely correct, but he’s also making exactly the same argument I am; that these people cannot actually achieve the technological revolution they are promising.
Zitron’s heavily focused on the economics. Specifically, he’s fixated on the cost of running the American data centers relative to their prospective future revenues and profit horizons. He’s not even “anti-Tech” relatively speaking. He’s just reading balance sheets and doing basic math on depreciation rate of hardware to conclude the current business models won’t work.
This isn’t to say it can’t happen. It’s to say these businesses won’t make it happen.
The problem is not how they’re pricing the outcome
This is where we’re in disagreement. They’re pricing their outcomes totally wrong, even in their self-proclaimed “best case scenario”. That’s leading them toward malinvestment - heavy spending on hardware and brute force algorithms. Marginal investment on material applications and workflow integrations that benefit anyone.
This isn’t a probability problem where they can luck into a multi-trillion dollar windfall.
Kudos for mentioning Xaiomi, they recently showed off their newest car factory, and it’s insanely highly automated, with workers basically only doing quality control!
China is probably the new masters of mass production, which I suppose is natural considering they have a home market of 1.4 billion people, they can invest in mass production for China only, and have a scale that is 4 times the scale of USA!
No country has ever had such a massive home market before China, and I think it’s a huge advantage for China regarding mass production.
No country has ever had such a massive home market before China
glances at India
But then that’s just the BRICS in a nutshell. Five countries with enormous internal consumer demand, which can afford to develop domestically to meet the needs of their local populations without worrying too much about export markets.
Again, we’re still in that “reading every over sentence” mode. For example, I say “hyper-advanced automation” and you reply claiming that “Good enough” automation is perfectly achievable. Yes. I know. I never said it wasn’t. I never said anything about good enough automation at all. And that kind of thing goes on throughout your response here.
By all means continue your conversation with whoever you think is making all these arguments, but they bare little resemblance to anything I’m saying, so there’s really no point in my responding any further.
Here “Good enough” matches the quality of most car brands and AFAIK surpasses Tesla.
To get quality like BMW or Mercedes, is probably only a matter of the automation being more expensive, there is no reason to think for instance Xiaomi couldn’t make higher quality automated assembly on more expensive cars.
Tesla assembly sucks, I have often heard reviewers complain about noises you wouldn’t find in way cheaper European cars.
I mean, I could point you to a few economics journalists who would argue otherwise. Ed Zitron’s been screaming about the downfall of AI for the last two years straight.
But the market can remain irrational longer than you can remain solvent, as the saying goes.
Doesn’t hurt to have a bunch of tech-friendly goobers running state and federal governments who are turning out the taxpayer’s wallet to finance these hallucinations.
I feel like you’re only reading every other sentence of what I say. In this instance, you seem to have fixated on this part, but sailed right past the part where I said that there’s zero evidence that anyone can actually produce hyper-advanced automation. I never argued that it was a rational decision to go all in on this possibility, and that’s entirely clear from my previous comments.
Ed Zitron is completely correct, but he’s also making exactly the same argument I am; that these people cannot actually achieve the technological revolution they are promising. That doesn’t change the fact that, if their wish granting genie was real, it would basically have unlimited upside. The problem is not how they’re pricing the outcome, the problem is how they’re evaluating the probability of achieving the outcome.
There’s plenty of evidence (Waymo, for instance, or Xaiome) that companies can produce “good enough” advanced automation. Tesla’s just not the guy to make it happen. Same with AI. Very possible to produce useful tools (Alibaba’s Deepseek, Insilico Medicine’s Pharma.AI, etc) with LLMs. Sam Altman’s just not doing it.
Zitron’s heavily focused on the economics. Specifically, he’s fixated on the cost of running the American data centers relative to their prospective future revenues and profit horizons. He’s not even “anti-Tech” relatively speaking. He’s just reading balance sheets and doing basic math on depreciation rate of hardware to conclude the current business models won’t work.
This isn’t to say it can’t happen. It’s to say these businesses won’t make it happen.
This is where we’re in disagreement. They’re pricing their outcomes totally wrong, even in their self-proclaimed “best case scenario”. That’s leading them toward malinvestment - heavy spending on hardware and brute force algorithms. Marginal investment on material applications and workflow integrations that benefit anyone.
This isn’t a probability problem where they can luck into a multi-trillion dollar windfall.
Kudos for mentioning Xaiomi, they recently showed off their newest car factory, and it’s insanely highly automated, with workers basically only doing quality control!
China is probably the new masters of mass production, which I suppose is natural considering they have a home market of 1.4 billion people, they can invest in mass production for China only, and have a scale that is 4 times the scale of USA!
No country has ever had such a massive home market before China, and I think it’s a huge advantage for China regarding mass production.
glances at India
But then that’s just the BRICS in a nutshell. Five countries with enormous internal consumer demand, which can afford to develop domestically to meet the needs of their local populations without worrying too much about export markets.
Again, we’re still in that “reading every over sentence” mode. For example, I say “hyper-advanced automation” and you reply claiming that “Good enough” automation is perfectly achievable. Yes. I know. I never said it wasn’t. I never said anything about good enough automation at all. And that kind of thing goes on throughout your response here.
By all means continue your conversation with whoever you think is making all these arguments, but they bare little resemblance to anything I’m saying, so there’s really no point in my responding any further.
Here “Good enough” matches the quality of most car brands and AFAIK surpasses Tesla.
To get quality like BMW or Mercedes, is probably only a matter of the automation being more expensive, there is no reason to think for instance Xiaomi couldn’t make higher quality automated assembly on more expensive cars.
Tesla assembly sucks, I have often heard reviewers complain about noises you wouldn’t find in way cheaper European cars.