It’s obvious to see why this is happening.
US AI is concentrated in a few super-large players with closed systems chasing AGI. Chinese AI is free to use by the world’s developers for today’s real world applications of AI. Chinese AI is even starting to become the default in the US among Silicon Valley start-ups.
Around the world, tens of thousands of businesses are going to want their own AI systems & they’ll build that on top of open-source foundations.
Interesting to see Chinese Open-Source AI ahead of the US in global usage figures.
It’s good news to see fossil fuel cars in such steep decline. This contrasts with the US, where gas cars are still near 90% of new sales. It’s China and Europe who are embracing EVs the fastest, though in China, combustion engine cars are still near 50% of sales.
But it’s not all good news. Half of those EVs are hybrid models. Data shows drivers with these still use a significant amount of gas, almost as much as ICE cars. The EU has set a 2035 deadline to ban all new gas car sales, and it seems that will include these polluting hybrid cars, too.
An irony here is I heard about this via an AI newsletter I subscribe to. It supports itself by selling ads for AI companies. The ad beside this story was AI-generated & touted a company called Genarena that sells AI-generated infographics. The example was a “Top 10 Liveable Global Cities”, where the No 1 & No 2 spot were the same city …
- still want to read books made by humans.*
True. LLMs just regurgitate. Not just that, they have no point of view or inner life. I don’t care what they think, in the way I might with James Joyce, Franz Kafka or Sylvia Plath.
That said, much of creative work is semi-automated already.
Is there really much true creativity in Hollywood superhero movies or TV soap operas? That kind of content seems like it could be largely done by AI.
New Pew research shows striking differences between AI experts’ views on AI and the general public’s views. On jobs, 21% of the public rate AI’s impact as positive, for AI experts, it’s 73%. On the economy, the corresponding figures are 21% and 69%. One of the few areas of convergence? Few on either side (9% & 11%) think AI will be good for elections.
Another interesting area of convergence? AI experts and the public mostly agree about which types of jobs are at risk from AI. One of the few differences is in self-driving AI. The public mostly thinks trucker jobs are safe from AI, but the experts don’t.
However, most of this polling reveal differences. Only 24% of the public think AI will personally benefit them. That figures is 76% for the AI experts. The big problem here? It’s the AI experts, and their interests, in the driving seat when it comes to legislation around AI.
The AI industry who has the US administration in its pocket. Their latest suggestion is to ban AI regulation at the state level, and only allow it at the federal level (which, conveniently, they completely control).
Even if they succeed, it won’t change the reality and public perception of AI, and it seems like it will store up conflict and clashes to come.
Pew Research - How the U.S. Public and AI Experts View Artificial Intelligence
At least it is starting to decrease in China - https://www.carbonbrief.org/analysis-chinas-co2-emissions-have-now-been-flat-or-falling-for-18-months/
Doctors in China have successfully treated a woman with type 1 diabetes using lab-grown insulin-producing cells made from her own tissue. Scientists reprogrammed her cells into stem cells, then grew them into small clusters capable of releasing insulin. One year after receiving the transplant, her blood sugar levels remain normal without any medication. This marks the first time in history that a person with type 1 diabetes has been cured of insulin dependence using cells derived entirely from their own body, not from a donor or embryo, paving the way for personalized treatments for millions.
This expands the range of ‘Work From Home’ to include physical labor. Humanoid robots aren’t far off the point (2030s?) where they can do most unskilled labor. With telepresence, they can take those jobs sooner.
This also brings something else closer. The looming crisis over what our governing economic model will be when human labor can no longer compete for wages with AI & robots.
Anyone who has ever read neurologist Oliver Sacks’ classic essay collection ‘The Man Who Mistook His Wife for a Hat’ might wonder about the downsides of having a protein nanowire brain extension. The lesson from the book is that small changes to the brain can have enormous consequences for consciousness and our experience of reality.
Who knows? Perhaps it might be like a permanent magic mushroom trip where you can see and talk to interdimensional machine elves, and that would be an upside for some people.
Yes, and there is also the possibility that it could be upon us quite suddenly. It may just take one fundamental breakthrough to make the leap from what we have currently to AGI, and once that breakthrough is achieved, AGI could arrive quite quickly. It may not be a linear process of improvement, where we reach the summit in many years.
Manna need to get quieter drones. People who live beside their base of operation don’t like the noise & disturbance.
Also, if you’re reasonably smart and self-motivated the 21st century world abounds with the materials to let you learn much of what you would in college. Not specialized learning maybe, but for generalized learning, yes.
An Irish drone delivery company Manna has been getting lots of complaints, apparently its not much fun living beside its base of operations.
https://www.rte.ie/news/business/2025/0820/1529313-drone-planning-dublin/
Won’t there be insurance for this?
If companies like FedEx can bear the cost of liabilities for huge numbers of human drivers, doesn’t that suggest the burden will be far less for robo-vehicle car companies?
Caveat - China Daily is owned and operated by the Chinese government/CCP. But the article is interesting in itself, and its official endorsement is interesting, too.
I’m still surprised at the rate LLMs make simple mistakes. I was recently using ChatGPT to research biographical details about James Joyce’s life, and it gave me several basic facts (places he lived & was educated at) at variance with what is clearly stated in the Wikipedia article about him.
I wonder will the US & EU bifurcate on AI adoption for government and administration, with the EU opting for open-source?
US models don’t seem interested in complying with EU law like the AI Act or GDPR.
If so, 5 or 10 years down the line this could lead to very fundamental differences in how the two territories are governed. There may all sorts of unexpected effects arising from this.
The person making this claim, Miles Brundage, has a distinguished background in AI policy research, including being head of Policy Research at OpenAI from 2018 to 24. Which is all the more reason to ask skeptical questions about claims like this.
What economists agree with this claim? (Where are citations/sources to back this claim?)
How will it come about politically? (Some countries are so polarised, they seem they’d prefer a civil war to anything as left-wing as UBI).
What would inflation be like if everyone had $10K UBI? (Would eggs be $1,000 a dozen?)
All the same, I’m glad he’s at least brave enough to seriously face what most won’t. It’s just such a shame, as economists won’t face this, we’re left to deal with source-light discussion that doesn’t rise much above anecdotes and opinions.
Former OpenAI researcher says a $10,000 monthly UBI will be ‘feasible’ with AI-enabled growth
thanks, I added it now.