Less artificial, more intelligent.
Our brains use only 20 watts of power to understand the world around us, and we often learn what we need to know from just a single example.
By that standard, today’s AI is profoundly underwhelming and painfully inefficient.
AI is using dangerously opaque algorithms that demand ever more data and computing power, even as it makes bewildering mistakes.
Consumed by memorizing the details of every tree, it never sees the forest.
Taut’s new approach is grounded in concepts from neuroscience, signal processing and information theory.
Reflecting the ways people reason and understand, it learns more while using less.
Our technology solves complex problems robustly, efficiently and transparently—just like the authentic intelligence that inspired it.
About us
Taut is a team of computer scientists, neuroscientists, mathematicians and engineers from Caltech, Berkeley, and UCLA.
We’re solving the most critical problems of today’s AI: robustness, security, interpretability and an insatiable appetite for training data and computing resources.
These problems are so fundamental that their solution requires an entirely new approach—one that will power the next generation of intelligent applications and devices.
