Chinese AI lab DeepSeek’s last model release was V3.2 (and V3.2 Speciale) last December. They just dropped the first of their hotly anticipated V4 series in the shape of two …
Gary Marcus has put forward articles theorizing that’s why the LLM/neural network models are so appealing to American capitalists. They at least have the appearance of something that can be infinitely scaled with investment (screw diminishing returns, right?)
Gary Marcus has put forward articles theorizing that’s why the LLM/neural network models are so appealing to American capitalists. They at least have the appearance of something that can be infinitely scaled with investment (screw diminishing returns, right?)
For sure, I think they genuinely think if AI got good enough they could just cull all the pesky workers at that point, and live like gods.