Shashank
103 posts


Please find this person and give them a raise.

In 2011, a neuroscientist at MIT named Dr. Li-Huei Tsai made a discovery that should have been on the front page of every newspaper on Earth. She exposed mice with advanced Alzheimer's disease to a flickering light pulsing at exactly 40 Hz — forty flashes per second. Nothing else. No drugs. No surgery. Just light at a specific frequency. Within one hour, the amyloid-beta plaques in their brains — the protein deposits that define Alzheimer's — began to dissolve. Not slow. Not gradually. Within sixty minutes. After seven days of daily 40 Hz exposure, plaque levels dropped by 50%. The mice regained memory function. Their neurons began firing in synchrony again. The brain's immune cells — microglia — activated and started clearing the toxic buildup like a cleaning crew that had been asleep for years. The study was published in Nature. The most prestigious scientific journal on the planet. Peer-reviewed. Replicated. Confirmed. That was 2016. It is now 2026. 40 million people worldwide have Alzheimer's. The pharmaceutical industry generates $13 billion per year from Alzheimer's drugs that do not reverse the disease. Not one of them. They slow it. Maybe. Temporarily. At $26,000 per year per patient. A 40 Hz light costs less than a dollar to produce. Dr. Tsai is still at MIT. Her research continues. Phase III human trials are underway. But you will not see this on the evening news. You will not hear your doctor mention it. You will not find it in any pharmacy. Because a frequency that costs nothing cannot sustain a $13 billion industry. The light is 40 Hz. The frequency is real. The science is published. And 40 million people are still waiting for permission to use it.

My dear front-end developers (and anyone who’s interested in the future of interfaces): I have crawled through depths of hell to bring you, for the foreseeable years, one of the more important foundational pieces of UI engineering (if not in implementation then certainly at least in concept): Fast, accurate and comprehensive userland text measurement algorithm in pure TypeScript, usable for laying out entire web pages without CSS, bypassing DOM measurements and reflow

JEPA are finally easy to train end-to-end without any tricks! Excited to introduce LeWorldModel: a stable, end-to-end JEPA that learns world models directly from pixels, no heuristics. 15M params, 1 GPU, and full planning <1 second. 📑: le-wm.github.io

This book is going to change how an entire generation thinks.













