My second short story release in English is ready: Tales of Illustrious Computer Scientists: Iola Varga, nun and computer scientist. invece.org/iola.html
That's undisputed, but the argument of the skepticals was a bit different: that even in the local space represented by the training set, LLMs could not interpolate on the realm of a deeply inferred representation, just on a shallow almost-already-seen one (parrot model).
Deepseek v4 flash 0731 running on M2 Ultra 192GB speed via @antirez's ds4 mxfp4, very usable, now we are pretty much token free.. will use it to try some interesting idea 🧙♂️
The anti LLMs folks continuously interpolate their skeptical vision with the current AI progresses. They were doing the biggest public perpetual error in the history of science. I don't care about Marcus (I focus on ideas that interest me) but this happened with actual top AI
Gary Marcus is right that math ≠ all of science.
But his bar just keeps moving:
2012: deep learning is "at best, only a small step toward the creation of truly intelligent machines"
2022: "deep learning is hitting a wall" months before ChatGPT
2023: fine, chatbots, but they
So what's your feeling on DeepSeek V4 Flash 0731?
Surely great model for Local AI and coding but clearly not at GLM 5.2 or Kimi K3 level. Obviously, because it's a different category.
For Frogger I tried 3 times using omp, but I always get a game too complex to play or with