If you’re looking for what’s next after deep learning and believe that reaching AGI requires exponentially improving the efficiency of current AI systems, you’re in the right place.
We're working on the next layer of the AI stack--
- A more reliable alternative to deep learning and LLMs using continuously trainable, compute-efficient weightless neural networks. We are building AI that can learn after training.
- We're talking AI that learns via real induction, with self-grounding, in a similar way to animals, because even though chickens aren't as "smart" as GPT, they don't hallucinate.
- many more things, join us on discord and visit our website for demos and more
