It’s not a bet, it’s just mania. There’s no rational basis for any of this, the entire thing is artificial. This impending disaster has more aligned with an oddly missing bubble in their diagram…the 2008 housing bubble.
I think we’re at the phase were people are trying to make wash their hands of culpability. That the bubble was not rammed down by a small group of people and was not entirely illogical when examined by more than 5 seconds of care, so that people can point to this in a little bit and go, “Nobody could see this coming!”
The real tragedy is not just the jobs and the economic destruction that we’re going to see, but the fact that this tech actually has some pretty cool use cases and it’s almost certainly going to get burned as when things like this collapse it just takes everything down with it.
open models won’t burn. local aficionados won’t stop using local. low cost low hardware requirements will become more of a focus as a push for embedded models that compartmentalize ai functions will be the boon. the markets may crash, but the knowledge and current hardware owned won’t
I am not suggesting that these don’t have value or won’t be used, but there are perfectly reasonable businesses that could have used these that are just going to get caught as collateral damage because LLMs are going to be a swear word in corporate planning for a while after all this.
There is always blowback during the collapse that is just as irrational as the bubble that caused it.
I’ve used LLMs dozens of times a day, everyday since about 3.5 and that’s just for my personal use cases not even including work. There is a weak argument to be made right now that coding is solved. Physical AI is exploding right now with robots getting tantalizingly close to actually being workers irl. What the hell are you talking about?
I've seen some interesting new robotics advances, but few have become viable products yet. An example that keeps coming up is laundry robots.
Here's one, https://www.weaverobotics.com/isaac-0, a $4000 static laundry folding robot. Of course it's only available in California and some unspecified percentage of folding will be completed via teleoperation.
Same company claims they will soon start selling an $8000 mobile maid bot that tidies and folds laundry.
IDK, it looks like it needs lots of work. If it was actually able to pick up my crap and fold my laundry, I'd be tempted, but it would also need to not phone home with camera data, the audio or other content of conversations. I'd also want an easy way to shut it down in an emergency.
When you can replace human sewing machine operators in garment factories, that's when you'll see widespread adoption of robots across the economy. Managing big pieces of fabric is hard.
I think we've still got a long way to go before we get jetsons style robot maids.
Unitree is not a joke. They’re pulling off some crazy stuff. That being said in the US Figure just put out like a nine hour video of it cleaning unseen homes. The tasks were all pretty basic, but it was able to do them and that’s just the start. Sunday robotics has a three hour stream of robots folding clothes correctly 99% of the time. We’re going to have a physical LLM lift off soon if we’re not already in it. They’re already seeing that increasing training data causes leaps in ability.
I have to stress this is not a real use case. It is a tech demo with no practical application. Doesn’t mean the tech might not have some application later in other places, but this is silly. The Unitree G1 costs $13,500.00 and does what? Clean a house? Show me actual demand and I’ll show you an actual market.
> that’s just the start.
Every single absurd AI claim can only speak about the future. We have invested more than the annual GDP of a Nordic country into this stuff in the last few years and have only forward thinking statements to show for it.
Yet the market can remain irrational for much longer than we can reason about.
The numbers are, as you (and Ed Zitron) point out, bananas, and the financial house of cards will collapse -- but there is enough fin-de-siecle chaos happening right now that anything is possible.
AI seems to be particularly useful in murdering people, so, for example, if there were another globe-wide hot war, it isn't clear to me that Altman et al don't virally weave themselves into the DNA of the state, both in the USA and in its adversaries.
In which case, the "bezzle" continues until the state collapses, whenever that is (depends who wins the war ;D)
The original sin of any modern forecast is proceeding from reason and sensibility. Now is the hour of the batshit.
> Yet the market can remain irrational for much longer than we can reason about.
It's already overdue.
> it isn't clear to me that Altman et al don't virally weave themselves into the DNA of the state, both in the USA and in its adversaries.
In the USA? Maybe. In its adversaries? Definitely no. Thanks to Trump's "Make America Negligible Again" campaign being extremely successful.
Adversaries and enemies alike, no exception, nobody trusts US any more, so everyone is already building their own stuff to get independent ASAP. Tech-bros have already lost their influence outside of the US.
Archive: http://archive.today/7RPo3
It’s not a bet, it’s just mania. There’s no rational basis for any of this, the entire thing is artificial. This impending disaster has more aligned with an oddly missing bubble in their diagram…the 2008 housing bubble.
I think we’re at the phase were people are trying to make wash their hands of culpability. That the bubble was not rammed down by a small group of people and was not entirely illogical when examined by more than 5 seconds of care, so that people can point to this in a little bit and go, “Nobody could see this coming!”
The real tragedy is not just the jobs and the economic destruction that we’re going to see, but the fact that this tech actually has some pretty cool use cases and it’s almost certainly going to get burned as when things like this collapse it just takes everything down with it.
open models won’t burn. local aficionados won’t stop using local. low cost low hardware requirements will become more of a focus as a push for embedded models that compartmentalize ai functions will be the boon. the markets may crash, but the knowledge and current hardware owned won’t
I am not suggesting that these don’t have value or won’t be used, but there are perfectly reasonable businesses that could have used these that are just going to get caught as collateral damage because LLMs are going to be a swear word in corporate planning for a while after all this.
There is always blowback during the collapse that is just as irrational as the bubble that caused it.
I’ve used LLMs dozens of times a day, everyday since about 3.5 and that’s just for my personal use cases not even including work. There is a weak argument to be made right now that coding is solved. Physical AI is exploding right now with robots getting tantalizingly close to actually being workers irl. What the hell are you talking about?
> Physical AI is exploding right now with robots getting tantalizingly close to actually being workers irl.
Citation needed, I have not seen any sign of this. I mean, China having humanoid robots running around is a cool demo and none of that is useful.
I've seen some interesting new robotics advances, but few have become viable products yet. An example that keeps coming up is laundry robots.
Here's one, https://www.weaverobotics.com/isaac-0, a $4000 static laundry folding robot. Of course it's only available in California and some unspecified percentage of folding will be completed via teleoperation.
Same company claims they will soon start selling an $8000 mobile maid bot that tidies and folds laundry.
IDK, it looks like it needs lots of work. If it was actually able to pick up my crap and fold my laundry, I'd be tempted, but it would also need to not phone home with camera data, the audio or other content of conversations. I'd also want an easy way to shut it down in an emergency.
When you can replace human sewing machine operators in garment factories, that's when you'll see widespread adoption of robots across the economy. Managing big pieces of fabric is hard.
I think we've still got a long way to go before we get jetsons style robot maids.
Unitree is not a joke. They’re pulling off some crazy stuff. That being said in the US Figure just put out like a nine hour video of it cleaning unseen homes. The tasks were all pretty basic, but it was able to do them and that’s just the start. Sunday robotics has a three hour stream of robots folding clothes correctly 99% of the time. We’re going to have a physical LLM lift off soon if we’re not already in it. They’re already seeing that increasing training data causes leaps in ability.
I have to stress this is not a real use case. It is a tech demo with no practical application. Doesn’t mean the tech might not have some application later in other places, but this is silly. The Unitree G1 costs $13,500.00 and does what? Clean a house? Show me actual demand and I’ll show you an actual market.
> that’s just the start.
Every single absurd AI claim can only speak about the future. We have invested more than the annual GDP of a Nordic country into this stuff in the last few years and have only forward thinking statements to show for it.
Yet the market can remain irrational for much longer than we can reason about.
The numbers are, as you (and Ed Zitron) point out, bananas, and the financial house of cards will collapse -- but there is enough fin-de-siecle chaos happening right now that anything is possible.
AI seems to be particularly useful in murdering people, so, for example, if there were another globe-wide hot war, it isn't clear to me that Altman et al don't virally weave themselves into the DNA of the state, both in the USA and in its adversaries.
In which case, the "bezzle" continues until the state collapses, whenever that is (depends who wins the war ;D)
The original sin of any modern forecast is proceeding from reason and sensibility. Now is the hour of the batshit.
> Yet the market can remain irrational for much longer than we can reason about.
It's already overdue.
> it isn't clear to me that Altman et al don't virally weave themselves into the DNA of the state, both in the USA and in its adversaries.
In the USA? Maybe. In its adversaries? Definitely no. Thanks to Trump's "Make America Negligible Again" campaign being extremely successful.
Adversaries and enemies alike, no exception, nobody trusts US any more, so everyone is already building their own stuff to get independent ASAP. Tech-bros have already lost their influence outside of the US.