[CuriouslyC]

If all you have is the starting point and the finishing point, the lack of the path taken from one point to another limits your ability to train models that can efficiently recreate the work, and increases its cost enough that it's possible the US labs can progress capabilities faster than Chinese labs can distill that behavior.

[lumost]

This just looks like a capex problem. There is no evidence that Anthropic has secret sauce above and beyond access to capital. If there is secret sauce, it's unclear that it changes the required amount of capital by all that much.

China will spend all of the money required to catch up, Google and OpenAI will both spend money to catch up as well. NVidia and others will not allow a frontier lab to become the AI bottleneck.

[drewda]

As of this month, everyone has 100+ pages from Microsoft on how they trained their MAI-Thinking-1 model: https://microsoft.ai/pdf/mai-thinking-1.pdf

OpenAI and Anthropic may have gone silent on how they build their models, but other companies have different incentives.

[sealeck]

> lack of the path taken from one point to another limits your ability to train models that can efficiently recreate the work

Isn’t this the problem inference (training) a model is designed to solve :)))

[jmalicki]

It is!

And it's a hard problem.

What's an easier form of training is being able to see the intermediate results and train to imitate them.

[wahnfrieden]

That’s already the case. Chinese ingenuity allowed them to achieve what they did without access to reasoning outputs

[throw10920]

This has got to be satire. Everyone, especially Singaporeans, know what "Chinese ingenuity" really is.

[wahnfrieden]

It’s merely descriptive of ingenuity required to distill models back into reasoning models without having any of the chain of thought. You underestimate the original work required because of biases

[throw10920]

It's a uniquely Chinese cultural feature to laude finding more clever ways to cheat and steal and deceive as "ingenuity".