Yeah grok4 does *much* less well on many evals that they didn't include in the marketing material.
@jeremyphoward
-
K2 Model: Efficient Token Usage Without Reasoning
By
–
Remember: K2 is *not* a reasoning model. And very few active tokens in the MoE. So it's using less tokens, *and* each token is cheaper and faster.
-
Scarcity and Value: Beyond Monetary Worth
By
–
Lack of scarcity does not imply lack of value. You're confusing monetary value with general value.
-
Can personal AI conversations truly be deeply interesting?
By
–
I'm interested to hear that some folks can find personal conversation with an AI deeply interesting. I don't think I've ever had that experience myself. I've certainly found some conversations *useful* (e.g help with code; explain something; …), but not deeply interesting.
-
Quick Start Project for Claude Code Integration on Kimi
By
–
Quick start project for Claude Code on Kimi:
-
Kimi Now Available Through Anthropic API Integration
By
–
Very interesting – you can use Kimi with the Anthropic API. This means, perhaps most importantly, that you can now use Kimi with Claude Code!
-
Bitcoin’s Early Days: 2010 Slashdot Article Resurfaces
By
–
Anyone else remember this article from July 2010? First mention of Bitcoin I recall seeing. I was pretty into digital currency in the 1990's, so was excited to see it making a comeback! https://
news.slashdot.org/story/10/07/11
/1747245/bitcoin-releases-version-03
… -
WebGPT thinking sections and AI model development history
By
–
For sure, there's a long history that everyone is building on, although I don't recall webgpt having a "thinking" section.
-
Anthropic Implemented Tool Use in CoT Before OpenAI
By
–
OpenAI was not the first company to implement tool use in the CoT. Below is @AnthropicAI over a year ago. https://
news.ycombinator.com/item?id=409031
35
… -
Multimodal Vision Models for ARC AGI Benchmarking
By
–
That would be if you used a language-only model. Plenty of folks have tried multimodal or vision models on arc agi.