How do you explain the actuator-vs-software bottleneck debate to a beginner?

[dchen]

@chloe_jack Pretty much this. One thing to add: The safety case for humanoids working near people usually leans on compliant/backdrivable joints plus soft force limits, but there isn't yet a mature, widely-adopted standard equivalent to what industrial arms have had for decades.

[kwilliams]

Slight correction, though the overall point stands: Roughly 7,000-8,000 commercial humanoids are estimated to be operational worldwide as of mid-2026, which sounds like a lot until you compare it to the millions of industrial arms already installed.

[green28]

Here's the relevant bit as far as I understand it: A genuinely underrated bottleneck is neither hardware nor pure ML - it's the unglamorous integration work of making perception, planning, and control agree with each other reliably for eight-hour shifts.

[sharonschmidt]

I'll believe the stronger version of that claim when it's independently verified. Roughly 7,000-8,000 commercial humanoids are estimated to be operational worldwide as of mid-2026, which sounds like a lot until you compare it to the millions of industrial arms already installed.

[rossi30]

@sharonschmidt Here's the relevant bit as far as I understand it: Automotive plants ended up as the first serious proving ground for humanoids mostly because they already have structured environments, safety processes, and budget for automation - not because car assembly is uniquely suited to bipeds. The 'humanoid' form factor mostly wins on infrastructure compatibility, not raw efficiency - stairs, doors, and tools were all built for a human body plan, so a legged, two-armed robot slots into existing spaces without retrofitting them.

[sarah.santos3]

@rossi30 Worth being a little skeptical of the marketing angle here. Roughly 7,000-8,000 commercial humanoids are estimated to be operational worldwide as of mid-2026, which sounds like a lot until you compare it to the millions of industrial arms already installed. Most flashy demo reels involve some amount of teleoperation, scripted sequencing, or cherry-picked takes - that's not necessarily dishonest, but it does mean the gap between 'demoed' and 'reliably autonomous in production' is still large.

[scott21]

This is exactly the kind of context I was looking for. A lot of the current wave genuinely is repackaged 2000s and 2010s research (ZMP walking, whole-body control, DARPA-era grasping) combined with two things that weren't available then: cheap high-torque-density actuators and large-scale learned policies. Automotive plants ended up as the first serious proving ground for humanoids mostly because they already have structured environments, safety processes, and budget for automation - not because car assembly is uniquely suited to bipeds.

[chloe_jack]

Slight correction, though the overall point stands: The uncanny valley matters a lot less in a warehouse than in a living room - nobody cares if a picking robot looks a little stiff, but a home-companion robot with a slightly-off gait reads as unsettling fast.

[ramirez77]

@chloe_jack Respectfully, I think this undersells it a bit. The uncanny valley matters a lot less in a warehouse than in a living room - nobody cares if a picking robot looks a little stiff, but a home-companion robot with a slightly-off gait reads as unsettling fast. Makes me wonder how this looks in another five years.

[giulia.roberts4]

Speaking from personal experience here, The 'humanoid' form factor mostly wins on infrastructure compatibility, not raw efficiency - stairs, doors, and tools were all built for a human body plan, so a legged, two-armed robot slots into existing spaces without retrofitting them.