September 29, 2026
Scan today's robotics headlines and you won't find a single new bipedal humanoid announcement. What you'll find instead is a robotic fish that scales up or down without losing its swimming ability, a soft robot that jumps forever on infrared light, and a profile of a roboticist who has spent his career building robots shaped nothing like humans at all. That absence is worth sitting with rather than skipping past, because it's a pattern, not a fluke -- this is at least the third day in a row we've had to note that humanoid news simply didn't show up.
Robots Guide's profile of Charlie Kemp, cofounder and CTO of Hello Robot, is the one story here with real business relevance, even though Stretch -- the robot he built -- isn't humanoid in the strict sense. It's a mobile arm on a wheeled base, designed around a simple question: what do people actually need help with at home or at work? Kemp's path from MIT AI research to building an assistive robot is a useful reminder that the most commercially durable robotics companies often aren't chasing the most photogenic form factor. They're chasing the task. For a business evaluating automation, the lesson isn't 'wait for humanoids to mature' -- it's 'figure out which narrow, repetitive task actually needs solving, and buy the machine built for that task.' We made a version of this argument in Robots That Help People Beat Robots That Impress Investors, and Kemp's career is a living example of it: usefulness beats spectacle, and it usually beats spectacle by years.
Robohub's report on EPFL's CREATE Lab and its scalable robotic fish, ScaFi, and New Atlas's piece on an infrared-powered jumping soft robot are both genuinely interesting engineering stories. The fish matters because propeller-driven underwater vehicles disturb the habitats they're sent to monitor, and a soft-bodied, scale-flexible swimmer solves a real environmental-monitoring problem. The jumping soft robot matters because it shows how untethered, light-powered actuation can replace batteries and motors in specific niches. But neither is a humanoid story, and neither should be forced into one. If your business is evaluating robotics for warehouse, retail, or field-service automation, these are worth filing away as signals of where soft robotics and biomimicry are heading -- not as something to act on this quarter.
Here's my honest take: the humanoid robotics narrative has been running well ahead of the news cycle that's supposed to support it. Investors and headlines have spent two years hyping general-purpose humanoids as the next platform shift, but the actual, fundable progress keeps showing up in narrower places -- an assistive mobile arm, an exoskeleton clearance, a scalable underwater robot. That's not a knock on humanoid research; it's likely that the hardware and control problems are simply harder than the demo videos suggest, and legitimate teams are working through them quietly rather than shipping vaporware. But if you're a business leader deciding where to put automation dollars today, the signal is clear: the ROI is in task-specific robots and, more often than not, in software that automates the decisions and workflows around those robots -- which is exactly the kind of operational plumbing tools like Workflow Automation or Internal Tools & Admin are built to handle right now, while you wait for the humanoid hardware story to actually mature.
So here's the question for you: if the humanoid form factor keeps missing headlines while task-specific robots like Stretch quietly rack up real deployments, how much longer should 'general-purpose humanoid' stay the default framing for what robotics automation even means?
Sources