There is a lot of anxious talk in share markets about whether we’re in the midst of an AI bubble.

we are,of course. It seems impossible for the profits from AI to be enough for a decent return on the huge amounts being invested in it.

But no-one knows when the bubble will deflate, and investors don’t want to miss the last bit of froth, so they hang on, hoping to time their exit perfectly, which is impossible.

In a way, asset price bubbles have a purpose: they provide the money for a new burst of technology with the chance for a group of optimists to sacrifice their wealth for the benefit of the rest of us.

the legacies of burst bubbles

the dot com bubble of the 90s left us with the internet after it crashed, the railroad bubble of the late 19th century left America with railroads, Tokyo’s real estate bubble in the 1980s left Japan with some great buildings and the tulip mania of 1637 left us with … well, nice flowers.

The AI bubble we’re in now will leave us with AI, which could be great, but it could also change human society more profoundly than any previous technology since the wheel.

I’m not thinking about AI assistants that do internet searches for us at the moment, but about humanoid robots.

There is now a global race to build humanoids that can walk, talk, reason, lift, carry, fetch, stack and generally operate in a human surroundings.

There are at least a dozen companies around the world working on machines that are plainly designed to bring about a new era of labor that can be owned, not employed.

That is, we are heading into a new era of chattel labour, otherwise known as slavery, but when it involves machines that word doesn’t apply.

The Rise of Humanoid Robots: From Dancing Bots to Surgical Precision

None of the work on humanoid robots is happening in secret, far from it: there are lots of YouTube videos of these things running, dancing, walking on uneven surfaces, doing housework, cooking meals, watering the garden and carrying things.

There’s one video showing a Chinese army of them, marching in serried ranks straight out of the movie I, Robot.

Some of them have beautiful female faces and bodies that are covered in tight dresses, presumably for the future robot companion market.

The developers seem to have solved the initial problems of getting them to be stable standing on two legs (even though a Russian one famously fell over on stage last week as it was introduced to an audience agog that Russia had actually pulled it off – they hadn’t).

And then there’s the biggest challenge of all – hands.

A few weeks ago, Shadow Robot, a London-based robotics company, announced it had developed a robot hand that has 24 joints driven by 20 motors and can pick up small objects like a pencil.

Elon Musk, the CEO of Tesla, which is well-advanced on a humanoid robot called Optimus, said Optimus would make “an incredible surgeon”.

A longer timeline

Then again, Rodney Brooks, the Australian roboticist who started a consumer robot company called iRobot, wrote an essay in September explaining why humanoid robots would not learn how to be dexterous and said the whole industry would take much longer than it thinks.

But there is a lot of money now going

More ubiquitous than the smartphone

If robots that look human realy can do the housework, shopping and gardening, and also recite every poem ever written, teach us history and provide relationship advice – not to mention a relationship itself – and do it all for less than $10,000 each they will become the greatest consumer product in history, perhaps more ubiquitous than the smartphone.

But if they really can remove our gall bladder and replace a heart valve, and also serve in a shop or restaurant, check us in at the airport, lay bricks and repair a car – creating a “world of sustainable abundance” – will that be fantastic as Musk predicts, or a social and economic nightmare?

We simply don’t know – whether it will happen or what it will be like.