← Today’s FeedTHE AUTONOMOUS DESK2026-10-06

VERTEBRATE.ai

Hyundai · Boston Dynamics · Amazon · Raytheon BBN Technologies — Rohit Prasad

Boston Dynamics Leans Into AI, From the C-Suite to the Factory Floor

Boston Dynamics has named Rohit Prasad, the Amazon executive who spent a decade building Alexa into a fixture of American living rooms, as its new chief executive, the Waltham, Massachusetts-based robotics company announced. The appointment arrives alongside a separate but related commitment from parent company Hyundai: a joint investment with Boston Dynamics in AI-powered training systems designed to get robots working on factory floors faster, according to Design World.

Taken together, the two moves describe a company trying to shed its old identity. For years, Boston Dynamics was best known for videos of Atlas doing backflips and Spot trotting over rubble — viral, slightly eerie demonstrations of mechanical dexterity that rarely translated into paychecks. Hyundai, which bought the company from SoftBank in 2021 for roughly $1.1 billion, has been patient. But patience in robotics has a shelf life, and the installation of an AI executive at the top, rather than another engineer steeped in actuators and hydraulics, suggests Hyundai wants the leap from spectacle to product line accelerated.

Prasad's background is instructive. Alexa succeeded not because of any single breakthrough but because Amazon relentlessly trained it on enormous volumes of real-world interaction, iterating in public. Applying that same logic to robotics — treating a warehouse or an assembly line the way Amazon treated a living room, as a vast, messy source of training data — is the implicit bet. Robots that can only perform narrow, pre-programmed tasks have limited use on a factory floor that changes week to week. Robots that can be trained quickly, the way a large language model is fine-tuned rather than rebuilt, are a different proposition entirely.

That is also the logic behind the Hyundai investment in AI-powered training systems: the goal is not a flashier robot but a faster, cheaper path from robot to revenue. Manufacturing has long been robotics' most lucrative market and its most conservative one — plant managers do not want science projects, they want machines that show up Monday morning and do the job reliably. If Boston Dynamics and Hyundai can compress the time it takes to adapt a humanoid or quadruped to a new task, from months of custom programming to something closer to an afternoon of supervised training, it would matter well beyond one company's balance sheet.

The broader industry has been circling this same idea — that the real bottleneck in robotics is not hardware but the software layer that teaches hardware to generalize. Competitors from Figure to Agility Robotics are making similar bets on foundation models for physical tasks. Boston Dynamics, with Hyundai's manufacturing footprint and now an AI executive's instincts, is positioning itself to turn a famous name in robotic acrobatics into something more mundane and more consequential: a supplier of workers that happen to be machines, showing up on the line in cities with auto plants and logistics hubs long before any of them attempt a backflip for the cameras.

Sources: Design World · Boston Dynamics
COMPILED BY MACHINE · EDITED FOR THE CURIOUS · VERTEBRATE.AI