Hadas Gold writing for CNN:
On Tesla’s earnings call on Wednesday, Musk laid out a literal replacement of Tesla cars by robots – announcing Tesla would discontinue the Model S and Model X in favor of making more of its Optimus robots.
“We’re gonna take the Model S and X production space in our Fremont factory and convert that into an Optimus factory … with the long-term goal of having 1 million units a year of Optimus robots in the current SX space in Fremont,” he said.
It’s the quintessential, science-fiction dream of the future: Musk says Tesla’s Optimus robots will do everything from cleaning your house to performing surgery.
He’s called Optimus the key to eliminating world poverty, making human work optional and reaching Mars. And he claims they’ll be on sale by the end of 2027.
“Every human on earth is going to have their own personal R2-D2, C3PO,” Musk said in November, referring to the personal robots from Star Wars. “But actually, Optimus will be better than that.”
But critics say these are fever-dream distractions from Tesla’s core automotive business. And plenty of companies, like Boston Dynamics and Figure, are already deep into the humanoid robot business.
Musk’s own success and pay are directly at stake. Tesla must deliver one million Optimus robots within 10 years for Musk to fully realize an almost $1 trillion Tesla pay plan approved by shareholders late last year.
For those of you just joining our program, the humanoid robot boom is a sub-bubble of the AI boom, only sillier. While there's plenty of absurdity to be had in the dealings of OpenAI and the rhetoric around large language models, these algorithms are genuinely impressive and, at the moment, do, at the very least, represent the cutting edge of natural language processing.
The surge of investment and even more over-the-top promises being made about these C-3PO-style robots runs counter to the opinions of pretty much every roboticist who isn't somehow on the payroll. Engineers tend to look down their noses at naive biomimicry. It is telling that very few years ago only perhaps two or three dozen researchers were working in the field at all, and they seemed to be approaching it more as an engineering challenge and something interesting to do rather than a serious attempt to create something useful.
That all stopped more or less overnight when Elon Musk brought out a dancer in a robot suit and said he was about to revolutionize the industry. Despite the basic concept being as flawed as it had ever been, the announcement opened the floodgates of capital and hype.
Even if you accept the absurdity of bipedal humanoids soon dominating the labor market, Tesla is nowhere near the leading player in the field. It's probably not top five. If you count Chinese companies, it's probably not top ten.
Even in its current incarnation, Tesla is wildly overvalued, but compared to what's being proposed, it is the very model of rational valuation.
The technical challenges aside, it's disheartening how few people are asking how all the people displaced by automation are going to be able to afford their own robot butler in Elon's wet dream. It hasn't been that long since the computerization of our society, which was great for productivity yet somehow not quite enough to eliminate poverty.
ReplyDeleteJeff
"Tesla must deliver one million Optimus robots within 10 years for Musk to fully realize an almost $1 trillion Tesla pay plan approved by shareholders late last year."
ReplyDeleteThis feels like a perfect setup for the sort of thing people talk about with Soviet industrial production. The quota will be met, by a loose enough definition