Naomi Bashkansky, a former OpenAI researcher specializing in AI safety and language model interpretability, left the lab in July 2026 to cofound Conduit, a San Francisco-based startup. The company is building AI models that decode brain activity into text without surgical implants, using electroencephalography. Conduit claims to possess the world’s largest neuro-language dataset, approximately 10,000 hours of recordings collected from thousands of participants in just six months. The goal is to create brain foundation models analogous to GPT architectures, capable of deciphering semantic content of thoughts rather than simply cursor movement intentions. This technology could transform communication for people with locked-in syndrome or conditions like amyotrophic lateral sclerosis.
Source: Read the original article
