The following is a comment on the Open letter to US President-elect Biden that was published by Asia Times on November 7.
It is obviously most timely to urge Joe Biden to put “rebuilding America’s pre-eminence in high technology” upfront, as a means to “build bipartisan support, capture the imagination of the country, and make your presidency one of historic importance.”
Asia Times “open letter” makes good points; but in my opinion, the vision of “technology” conveyed by the admittedly brief text is woefully inadequate.
Except for mentioning “materials science, fission and fusion” in one place, the bulk of the letter places what I consider to be a fatally one-sided emphasis on chips, telecommunications and artificial intelligence (AI).
Permit me a bit of sarcasm in some of my comments – maybe a bit unfair to the authors of the letter, but for the purpose of adding a point that it fails to address adequately.
Semiconductor manufacturing and high-speed data transmission technologies are strategic issues for the US, no quarrel with that. They are relevant to national security. But people may be tempted to look upon AI and digital technologies – including chip manufacture – as the deciding factors in the battle for technological pre-eminence. In my view that would be a very big mistake.
AI is obviously very important, although it suffers from excessive hype and often misplaced emphasis. The key issue is: Who will be in a position to use artificial intelligence in the most intelligent, broadest and most productive way? That depends on economic factors that have nothing to do with AI per se.
Despite high-profile statements and programs, the Chinese are not so stupid as to focus one-sidedly on the mentioned areas. They are building up in-depth capabilities across the board, in all advanced areas of science and industry. Sometimes upfront, often much more quietly.
If people think chips + 5G + AI technology is the decisive battleground, they are in for some nasty surprises.
Moreover, let’s suppose the US retakes world microchip production and digital technologies in general. Hurrah! Won’t it be great to have US microchips and AI software installed in European, Chinese, South Korean and Japanese products? In their advanced manufacturing equipment, automobiles, aircraft, high-speed rail, power plants, urban transport systems, construction machinery, etc that would be dominating the world market?
The letter fails to place the issue of technological pre-eminence even minimally into its economic context.
For example, the word “infrastructure” does not appear even once in the open letter, although it could be a tremendous bipartisan issue for the coming president.
Hardly anybody seems intelligent enough to link the infrastructure issue to that of “technological pre-eminence.” Are we going to rebuild our basic infrastructure with 1950s technology? Or with 2050s technology?
Infrastructure includes high-speed data communication, of course, but there is a lot more, to put it mildly. Rebuilding infrastructure means more than installing 5G (fifth-generation) transmitters on old telephone poles.
Imagine the magnitude of market demand that rebuilding America’s basic infrastructure – with 21st-century technology – will create for the advanced manufacturing sector. Including for AI, and including things like the utilization of novel, nanotechnology-engineered materials on gigantic scale, perhaps comparable even to that of steel today.
One often hears of a $2 trillion “infrastructure funding gap” in the United States. But that figure is ridiculously small compared with the investment you will actually need to rebuild US infrastructure on a 21st-century level, as opposed to just filling potholes and replacing rusty bridges.
According to sector estimates, rebuilding America’s ancient electricity grid alone would cost around $5 trillion. And what if we get serious about the so-called hydrogen economy, and/or a complete electrification of ground vehicle transport? And what if we get serious about eliminating America’s dependence on fossil fuels for the vast majority of its base-load electric power generation?
Unfortunately people still associate basic infrastructure with digging holes and pouring cement. Dirty and sweaty work, they think; we can get immigrant labor from Mexico to do that.
It would be better, of course, to use robotic construction equipment, already beginning to come online. But the way things are going, this will have to be imported from China. We might console ourselves with the thought that the microchips and maybe even part of the data transmission system might be US-made.
And where is the US high-speed rail system? China’s 40,000-kilometer-long high-speed rail system is the pride of that nation, and Japan is building 500km/h maglev. High-speed rail in the US is dead in its tracks. A joke. A disgrace.
It’s great that we have a Hyperloop test bed. Will the US build Hyperloop systems on a scale that will make a dent in passenger transport? Or will Hyperloop go the way of hundreds, maybe thousands of revolutionary technologies that American scientists and engineers have created – technologies that worked, but which never made it beyond the testbed stage?
(The molten salt fission reactor is just one example that comes to mind. Who developed the technology? US scientists and engineers. Who it is building it now? China! With improvements, naturally.)
Much or most of the US rail-track network is 1930s vintage, or older. We can learn from the Chinese how to lay rails quickly, and with great precision. And then maybe our trains will be able to run at more than 30km/h.
The Chinese, Japanese and South Koreans can also help with urban infrastructure. The New York subway and elevated train system is famous for its 1930s equipment. Nearly a century later, the stuff belongs in a museum.
You might make great chips and fantastic 5G communication networks, but not be able to drive anywhere because of the potholes in US roads.
You can still fantasize about “smart cities,” using big data and AI to keep the 1930s equipment running for another decade.
Thanks to artificial intelligence you might even get autos to jump automatically over the potholes, so that you wouldn’t have to fill them. (Unfortunately, not a few AI applications in the US nowadays have a bit of that character: adapting to problems that shouldn’t have existed in the first place.)
The world semiconductor market is roughly $300 billion. That’s fairly substantial, but just a third more than the $200 billion commercial aircraft market, which we Americans are in the process of losing to Airbus and – yes, they are coming fast – the Chinese.
The world automobile market is 20 times as large as for semiconductors – around $6,000 billion. The US share has collapsed to around 13%. Thankfully, half of American drivers still buy American cars. Presumably for sentimental reasons.
No problem, someone will say, as long as we make sure the foreign cars use American microchips, and maybe US-made AI software for autonomous driving. (The sensors will come from China.) If we are pre-eminent in those fields, then why should we care who manufactures the rest of the vehicle?
Exactly the kind of thinking that led to the catastrophic erosion of the US industrial base in the first place!
Leaving black humor aside, I would suggest that the president-elect launch a bipartisan initiative that would address both issues at the same time: regaining technological pre-eminence, and rebuilding America’s infrastructure on a 21st-century level.
We could toss in a serious US space program – which doesn’t exist at the moment – a serious fusion program and a few other things that would “capture the imagination of the country” a lot more than computer chips, fast data links and talking robots.
Jonathan Tennenbaum received his PhD in mathematics from the University of California in 1973 at age 22. Also a physicist, linguist and pianist, he is a former editor of FUSION magazine. He lives in Berlin and travels frequently to Asia and elsewhere, consulting on economics, science and technology.