Kris Hansen

Yann LeCun and Randall Balestriero just Proved the Platonic Representation Hypothesis, and found a better way to achieve it. The Platonic Representation Hypothesis suggests that AI models trained on different modalities, like images and text, eventually learn similar representations of the world. But this paper shows that how models are trained matters more than simply scaling them up. So they mathematically show that, under certain conditions, two JEPAs trained completely separately can learn essentially the same underlying representation of the world. One learns from different images of the same scene, while the other learns from different descriptions of that scene, without ever being trained on image-text pairs. Their approach achieves the same image-text alignment with 42x fewer training samples than next-token prediction. A 109M-parameter text JEPA even outperforms LLMs up to 72B parameters on cross-modal alignment benchmarks. alphaxiv.org/abs/2610.39302

The new 5-inch Macbook Pro has touchscreen and 5G.

1/4 Meet Qwenputer: Qwen3.5-4B with a full RISC-V CPU inside its weights, using the model's own Gated DeltaNet layers as RAM. It runs C and MicroPython in its forward passes, no external runtime. On DeepMind math: base 55–90%, base + Python tool 95–100%, Qwenputer 97.5–100%.

Had to run Mbappes El Clasico winner from this week through the spatial pipe. Not too bad for one broadcast shot and all local models. ⚽️💯🤯Shot on @Apple Vision Pro. #vr #barca #soccer #ar @realmadrid @Barca

Canada continues to get industrial policy wrong. Picking winners creates artificial barriers to competition and misallocates capital. A better approach would be a broad-based subsidy accessible to every firm in the space.

I’ve survived Perl, SOAP, XML, the microservices hype cycle, and four distinct eras of JavaScript frameworks. You cannot intimidate me with a new build tool.