Inside the rise of Ashok Elluswamy, the Elon Musk lieutenant leading Tesla's self-driving gamble (www.businessinsider.com)

Ashok Elluswamy, a long‑time Tesla Autopilot engineer, has quietly become the company’s chief executor of Elon Musk’s highest‑risk bets: he led the 2022 switch to a camera‑only perception stack (removing ultrasonic sensors), oversaw the public Full Self‑Driving rollout, and in 2024 expanded responsibilities to include Tesla’s robotaxi pilot, ride‑hailing launches, Optimus robotics work, and parts of the in‑house chip team. Colleagues describe hands‑on testing—engineers deliberately testing near‑miss scenarios—and big organizational shifts under his watch, including prioritizing large‑scale data collection over traditional coding roles and trimming staff to match the new AI‑first strategy. Technically, Tesla’s pivot is a live experiment in scaling end‑to‑end, vision‑based autonomy without sensor fusion, relying on fleet data and neural networks to handle perception and prediction. That approach accelerates learning from real driving data but raises ML concerns: dataset bias (reports that training prioritized routes Musk used), overfitting to VIP scenarios, validation gaps in safety‑critical systems, and contentious demo practices that have drawn legal scrutiny. For the AI/ML community, Elluswamy’s work is significant because it tests the limits of camera‑only models at production scale, informing debates on sensor fusion vs. vision‑only architectures, data curation, and how to verify and validate large, safety‑critical autonomous systems.