Comment by Sanzig
8 hours ago
Agreed that you can export-control the closed source models pretty well (although I think the administration is gravely underestimating the long-term damage that will do to the US economy).
The bigger problem will be if someone (such as a Chinese AI lab) releases a Fable/Mythos-class open weight model. That you can't really export control successfully. Sure, you could class it under EAR or ITAR, but that's just going to make using it difficult for American companies, not everyone else. It would be a stupid protectionist measure that would only hurt the US - so I fully anticipate the admin would try it.
I can't think of anything better that could happen than a Mythos/Fable level model released with open weights. It would be a huge step forward for the whole world.
Particularly if it came paired with DeepSeek pricing.
Alternatively, we could try to keep closing barn doors after the horses are already in the next county, have been hit by a car, and are on their way to the rendering plant.
Oh, I wholeheartedly agree! My "bigger problem" statement can be read as "a bigger problem for the US government."
>The bigger problem will be if someone (such as a Chinese AI lab) releases a Fable/Mythos-class open weight model.
That is my dream for 2026 right now!
None of your statements follow eachother. Anthropic did a few premature optimizations so they could immediately show their ineptitude. If it wasn't their constant advertising (safety focused lab whose model escapes their sandbox every other week with an unlisted exploit, seem to leak models or proprietary code every other week). It would be their insane pricing for a basic utility, intentionally locking the rest of the economy (And most major corporations) out of the (minimal, cos it's AI not electricity) gains. Of course there was never actually a mote. It's a freaking machine learning model. GLM5.2 was trained on a tenth the hardware Mythos was trained on. If anything Anthropic are already 6 months BEHIND. They are BEHIND google in terms of true multimodal intelligence as well. They are ahead of their partner in crime OpenAI.
Isn't GLM5.2 already Mythos class?
How do you qualify what makes a model "Mythos class", and how do you reliably test for it?
Presumably a deepswe benchmark, which IIRC puts GLM 5.2 between opus 4.8 and fable.
Opus and Codex are both catching plenty of very good improvements to my GLM plans. It gets a lot right too, has a lot of good things it does, good habits and practices. But it's not as smart, not as observant, not as able to craft a nice system. In my experience.