What if you could squeeze way more signal out of a brain-computer interface (BCI), *without* scientific or engineering breakthroughs on the BCI hardware / sensors? Well, you can. With data. *Lots* of data. This is the emerging field of neuroAI: neuroAI = AI scaling laws x BCI It's the bitter lesson for the brain. neuroAI is unlocking the ability to decode thought-to-text, thought-to-image, and more at capability levels that even science fiction scarcely imagined. (I can vouch for this from the field -- I'm knee deep in it myself.) neuroAI's on track to revolutionize neuroscience, human-computer interactions, and even what it means to be human itself. It's bci/acc. neuroAI will also profoundly impact frontier AI. Imagine RLHF, but without slow humans in the loop. That's just for starters. There are a dozen different ways neuroAI will catalyze AI. My friend Sat just published a thoughtful, well-researched piece about neuroAI and where it's heading. It's the first thorough synthesis across fields that I've seen. Truly 💯. It's linked below. I already view Sat's piece as essential reading for anyone wanting to understand neuroAI, and how it relates to AI, data, neuroscience, BCI and more. Enjoy:) Great stuff Sat!! 💪

You can grow up in the middle of nowhere, yet dare to chase your biggest dreams. In my case, I grew up on a farm in rural Canada, through the dustbowl of the 80s, with long hours of backbreaking work. Getting a TRS80 and learning to program was an inflection point. From that, I took every chance to ratchet up, towards my AI dreams. This podcast zooms in on my early years, from the farm, to university, to doing AI at National Defense, to my first AI startup, and beyond. Thanks @Mihaela_Ulieru for the fun conversation! :)