nphard85

The alternative is to believe that the mind, conditioned and optimized by millions of years of evolution for the survival of the organism under constant threat of hostile natural forces, can capably perceive reality as it is.

Could the (not so perfect but technically simple) solution be to transform the style of content under each tag to the correct expected style for the tag, via a smaller or purpose-built LLM, before the data stream is fed into the main LLM? Perhaps the two LLMs can be co-trained to keep the overall quality of the output stable while role confusion is minimized.

exactly

Will there be refund?

You sound bitter. May be you don't realize that.

To be fair, the author does mention the huge difference between Gemma 3 and Gemma 4 on Tau function calling benchmark.

Very interesting. How does this approach work for complex agentic workflows where the LLM is expected to orchestrate across multiple tools (such as when using MCP)? Or is this mainly for simple cases like the ones presented in the blog post?

Dwarkesh*

The current CEO is Sridhar Ramaswamy

Interesting. Is there a source for this claim?