Hopefully once capitalism implodes, we’ll be able to move to a more sane world.

I’d argue that we’re just starting to reach that moment right now. Cause a lot of tech people are now using Chinese LLMs and realizing they’re not actually behind. And it’s causing a lot of existential angst right now.

indeed

:)

yeah this is fantastic stuff

the paper https://www.nature.com/articles/s43588-026-00956-4

my favorite part is that it’s so obvious how unhinged that comment is and he’s so out of touch that he didn’t realize it

indeed

it’s too bad it doesn’t appear to be maintained anymore, we really need an active project like this

yeah it’s pretty impressive

it would be pretty awesome if we had completely open models that are community trained, there’s already this too https://github.com/bigscience-workshop/petals so probably could be updated and modernized

It’s kind of similar to what you see with Fable incidentally. They seem to be very token heavy, so the effective way to use it is to get it to do high level investigations and then farm out the actual implementation work to something like GLM.

I want this

Yeah, the whole business model is falling apart. They can’t monopolize the market, they don’t have a better product, and their input costs are 3x of those than companies in China. It’s a complete disaster. They got high on their own supply chasing their technological singularity while Chinese companies pragmatically took the slow and steady approach without burning through insane amounts of capital.

Anthropic is basically fucked now. Nobody in their right mind will pay for Opus 4.8 when Kimi is a fraction of the price and performs closer to Fable. So, the whole idea of keeping Fable exclusive is now dead in the water. But I think the very fact they felt the need to do it tells us something as well. Before, they’re just release every model as a regular upgrade for all users, but with Fable they wanted to only make it available on pay to play basis. To me this suggests that either it costs a fuck ton of money to run cause it’s inefficient as fuck even compared to their current models, or they’re starting to hit the ceiling of what they can do with the architecture and not expecting to put out anything significantly better in the near future. Either way, Kimi just pulled the rug from under them here.

I don’t think there’s anything that fundamentally prevents AGI from being developed, but I strongly suspect that it’s going to take more than LLMs to do that. In fact, frontier is already moving towards world models where a physics simulation becomes the basis for reasoning. Deepseek gunning for AGI is not in any way at odds with the current methods hitting a plateau. We already know that simply making models bigger doesn’t scale. It’s also unlikely that feeding more data into models is going to help either. The problem is with how the architecture works, and making any substantial leaps from here will require rethinking that.

And putting AI in charge of R&D is tangential to improvement of models themselves. Current capabilities are already very useful, and when you put a model in a feedback loop, it can indeed work autonomously and produce useful results. So, current crop of models will absolutely lead to a ton of breakthroughs and new types of automation that weren’t possible before. China is also very well positioned to take advantage of all that being the factory of the world.

The question is what it will take to make models themselves significantly more capable though. Look at the progress that was being made like two to three years ago. Every few months we’d see some incredible breakthrough in capability. That’s basically stopped in the past year, and progress has become very incremental now.

Yup, and also notable how China proceeded to invest in energy infrastructure before trying to scale out compute. The Americans put the cart before the horse here.

They just keep getting high on their own supply. While American companies are betting on the idea that if one model can pull away it’s going to keep self improving and nobody will catch up, Chinese companies are betting there will be a plateau to this tech, and it’s more important to focus on market dominance because they will catch up later.

And it’s becoming clear that Chinese companies made the right bet because Chinese models are closing the gap now, which means we are getting to the point of diminishing returns already where it takes ever more effort into squeezing just a bit more capability on the frontier.

For sure, as China keeps pulling ahead we’ll see world standards gravitate towards Chinese tech, and the west will start becoming a hermit kingdom.