In an interview with Bill Gates, Sam Altman says he is super-excited for AI applied to robotics.
Sam Altman – “We have to go do this thing. This is now an unstoppable technological course. The value is too great.”
Tens of billions of dollars are going into the effort to dominate technology and transform the world with Large Language Models and AI.
The scale of effort will grow and reshape society even more than the industrial revolution, the early adoption of computers, smartphones or the internet.
OpenAI has started to buy some robotic hardware companies.
The humanoid robot revolution to replace, enhance and augment everything associated with labor is critical to the full realization of a transformed civilization.
There are some forward thinking analysts who look at Tesla or other great technology companies as the sum of the parts of future businesses. When they analyze Tesla they always realize that FSD/robotaxi and humanoid robot Teslabot have the largest potential. The analysts then discount and reduce the potential revenue. They give more weight to the mundane and established lines of business. However, IF it is certain that advanced AI and humanoid robotics can be solved and perhaps fully solved within two to five years, then the choices in what to prioritize is changed. Winning humanoid bot to the point where it can multiply factory productivity and enable replication of factories is the final prize. It would be like Henry Ford creating the 10X faster assembly line but then also continually 3X productivity every year.
IF there is a clear path to fully realizing AI at AGI and ASI levels.
It is could be “just” properly spending a few hundred billion dollars to scale the AI training issues and rapidly improve the AI and AI for humanoid robots.
Microsoft and OpenAI have spent about $10 billion to develop GPT4 and soon GPT5.
Google has matched GPT4 with Gemini and is keeping pace in the competition with GPT5.
Dozens of companies have collectively spent about $150 billion to develop robotaxi and perfect self driving cars. There is revenue and value from advanced driver assistance systems. Large Language Models are showing they can go beyond human expert capability.
There is evidence to suggest that AI can match or go beyond human expert capability for driving and labor. LLM and neural nets needed many innovations to achieve greater than human expert capability.
There will not be a shortage of resources to test the hypothesis that full human replacement humanoid bot can be developed.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
I doubt the cost-benefit ratio, including maintenance & charging costs, will make sense for the average home for 5-10 years. It might replace some toys but, see the movie Megan or others for the horror end of things.
It could transition Amazon employees out of floor work, maybe out of warehouses altogether. That’s already pretty automated and it’s not clear that humanoid robots are actually the best approach, e.g. why have a robot that can climb a ladder when you can program an AI forklift with graspers to retrieve items off a top shelf to deliver to a packing machine, then to a conveyor belt machine to load into a FSD truck. No more heating/cooling warehouses, no more lighting when using RFID & geolocation, certainly no more DEI/EEOC, harassment or discrimination cases.
Of course, when labor is displaced at scale, there’s going to be mass social unrest, maybe even armed robot bashers, so some smart programmer egged on by cost-conscious owners will program self-preservation into robots that go out in public (already, those self-propelled delivery carts experimentally used in Charlotte, NC and Phoenix, AZ to deliver morning meals are subject to kicking over, into ditches, etc.). Then, it’s just a small AI leap to taking defensive actions, maybe calling in defender robots, backed by property rights laws. Of course, when it comes to choosing, a jobless human is still more valuable than a robot worker who works tirelessly without complaining – except for self-preservation alerts to the owners – 24/7, so there will be safeguards against “excessive resistance,” right? Right? …Right?
Unless society comes up with path to treating AI along with all IP as a sort of “Land” in the Georgist sense, that is residually owned by everyone with just a temporary tranche of IP property rights for creators – the economy will break. This is obvious, fair and moral. AI is just the distilled common intellectual heritage and AI will create most of future IP based on that. AI and IP rents have to be owned in equal and inalienable shares by everyone. That still allows plenty of private property and inequality to drive economic competition – it just eliminates the extreme version.
you will own nothing and you will be happy seems to be the scary sentiment here
When the WEF and Klaus Schwab own nothing, *then* we can be happy. 😉
IMO it depends on the speed of software progression. I think that if humanoid robot will cost 10-20k and the software is good enough, many people will buy it for home usage.
I think that there are 2 main issues with humanoid robots:
Firstly, people continue to underestimate how complicated most of not all “simple jobs” are. It’s almost like people went from high school to college to high-skilled jobs and never worked a “simple job” in their lives or it was so long ago they completely forget what it is like. Even many humans are mediocre at “simple jobs” and they have had millions of years to become proficient at adapting, learning, applying dexterity to new tasks, etc.
Secondly, many of the jobs people are envisioning humanoid robots doing are ones that can face competition from legless robots using the same “AI” as the humanoids. If the ability to climb stairs and step over trash was a universal job requirement there would be no people in wheelchairs working anywhere.
I’m not saying there won’t be some role for humanoid robots but it won’t be soon and they won’t corner the entire labour market.
I think you miss the point. Humanoid robots won’t replace specialized robots, they will fill in gaps – it’s just that those gaps now represent most human labor. For example the same inference engine boards like HW4 that run Optimus bots and Tesla vehicles could run industrial machines in Tesla “Alien Dreadnought” GFs and work vehicle robots or other specialized robots. Work vehicle robots are versions of Fork Lifts, Combines, Garbage Trucks, Bulldozers, Street Sweepers, Snow Plows, etc. It would be possible to train an Optimus to sit in a conventional version and operate it – but it’s obviously less efficient ultimately than just putting the Brain in the vehicle.
The point about humanoid bots is that their universality makes demand for them effectively unlimited for the foreseeable future so genuine mass production by the hundreds of millions is easily justified. Robotic lawnmowers are better than Optimus + lawnmower and will eventually fill that niche but it is just a limited niche.
The point is not that there aren’t a lot of tasks that humans will be better at for awhile but that there aren’t enough to slow the replacement of human labor.
Think about B2B for AI robots. Factory, warehouse, delivery.
Any job that costs $100k to fill and supervise ($50k employee pay) and has several employees with similar repetitive tasks.
AI for home is further off because most people don’t currently pay someone to do household tasks.