The economic, societal and political implications of Physical intelligence - or autonomous labor - will be even more profoundly disruptive to societies and economies than AI, and 'leaders' are completely unprepared to deal with this. The fundamental changes that robotics will bring are going to shake so many of our foundational structures to their core and make us question a lot of beliefs and policy frameworks that are 'settled'.
Take demographics. For at least 20 years, leaders in Europe and the US have taken it as established truth that immigration is necessary to counter falling demographics. That because of falling birth and worker-to-retiree ratio's, it is absolutely necessary to import people - or 'labour' - to be able to maintain economic growth and fund the social services costs of growing retiree populations. But with autonomous labor, this truism is not only wrong, it is completely 180 degrees wrong. In this new era of mass job replacement by robot, immigrants just add to the problem of what to do with suddenly 'surplus' labor. This is obviously going to be absolutely explosive politically and socially.
What is the social, political, economic and demographic structure that benefits most from this brave new world? The ideal (practically, not necessarily morally) characteristics of a country in robot world is that it has cheap energy, lots of capital (to invest in new zero human robo factories), good logistics, very little legacy demographics that have to be supported with social services, and it has - or can attract - the smartest, most high value adding people to create, engineer and manage the new companies that emerge.
That basically describes the UAE. Think about it in contrast to the legacy economic powerhouses the US, Germany, Japan and China. Those countries may have a huge problem on their hand with the combination of aging populations (expensive social services), and large working age populations whose services are suddenly no longer nearly as much in demand. Plus Germany and the US have the added problem of increasing tensions with immigration.
Contrast that with the UAE. Its industrial development model has been to import enormous numbers of foreign workers to build all of its infrastructure, and then send them home when they are done. No legacy social services costs, no mass of newly underemployed (and angry) population to deal with. It has the cheap energy and basically unlimited investment capital to massively build out and operate a new robo industrial zone, and with zero income taxes it is already attracting high value professionals from high tax locations. It also has a modern logistics network that gives it advantages globally, and can allow it to totally dominate regionally (assuming the Iran situation is sorted).
I had dinner the other night with a Frenchman (some of you may know him as PEG) who condensed some of your very penetrating analysis with the saying “better WALL-E than Wali.” That is, he welcomed the AI revolution on the grounds that it destroys the rationale for mass immigration. I can see the logic of that, but it seems like a pretty dark solution.
In brief (best of my understanding, not my specialty), it's an alternate architecture to VLM being pursued by Yann LeCun. VLMs work by associating images with natural language. But JEPA does not - JEPA cuts out the language layer and trains on the association of images with their lower-level mathematical representations. In this way, LeCun intends to create AI that reacts to changes in the world, instead of reacting to changes in the way we write about the world.
The economic, societal and political implications of Physical intelligence - or autonomous labor - will be even more profoundly disruptive to societies and economies than AI, and 'leaders' are completely unprepared to deal with this. The fundamental changes that robotics will bring are going to shake so many of our foundational structures to their core and make us question a lot of beliefs and policy frameworks that are 'settled'.
Take demographics. For at least 20 years, leaders in Europe and the US have taken it as established truth that immigration is necessary to counter falling demographics. That because of falling birth and worker-to-retiree ratio's, it is absolutely necessary to import people - or 'labour' - to be able to maintain economic growth and fund the social services costs of growing retiree populations. But with autonomous labor, this truism is not only wrong, it is completely 180 degrees wrong. In this new era of mass job replacement by robot, immigrants just add to the problem of what to do with suddenly 'surplus' labor. This is obviously going to be absolutely explosive politically and socially.
What is the social, political, economic and demographic structure that benefits most from this brave new world? The ideal (practically, not necessarily morally) characteristics of a country in robot world is that it has cheap energy, lots of capital (to invest in new zero human robo factories), good logistics, very little legacy demographics that have to be supported with social services, and it has - or can attract - the smartest, most high value adding people to create, engineer and manage the new companies that emerge.
That basically describes the UAE. Think about it in contrast to the legacy economic powerhouses the US, Germany, Japan and China. Those countries may have a huge problem on their hand with the combination of aging populations (expensive social services), and large working age populations whose services are suddenly no longer nearly as much in demand. Plus Germany and the US have the added problem of increasing tensions with immigration.
Contrast that with the UAE. Its industrial development model has been to import enormous numbers of foreign workers to build all of its infrastructure, and then send them home when they are done. No legacy social services costs, no mass of newly underemployed (and angry) population to deal with. It has the cheap energy and basically unlimited investment capital to massively build out and operate a new robo industrial zone, and with zero income taxes it is already attracting high value professionals from high tax locations. It also has a modern logistics network that gives it advantages globally, and can allow it to totally dominate regionally (assuming the Iran situation is sorted).
It is going to be a wild ride...
I had dinner the other night with a Frenchman (some of you may know him as PEG) who condensed some of your very penetrating analysis with the saying “better WALL-E than Wali.” That is, he welcomed the AI revolution on the grounds that it destroys the rationale for mass immigration. I can see the logic of that, but it seems like a pretty dark solution.
As you say, it’s going be a wild ride.
Matt, you should know about JEPA.
In brief (best of my understanding, not my specialty), it's an alternate architecture to VLM being pursued by Yann LeCun. VLMs work by associating images with natural language. But JEPA does not - JEPA cuts out the language layer and trains on the association of images with their lower-level mathematical representations. In this way, LeCun intends to create AI that reacts to changes in the world, instead of reacting to changes in the way we write about the world.
This is a great video series on this: https://www.youtube.com/watch?v=kYkIdXwW2AE
If I was teaching my robot how to move, I would just make it watch Crouching Tiger Hidden Dragon on a continuous loop. Then, watch out!
Woohoo! New Crawford drop!