Our tech overlords tell us that artificial intelligence (AI) will soon create a world of abundance while solving our major problems such as climate change. Just so you know who I'm talking about, I have previously defined our tech overlords as "a grab bag technology companies, technology scientists and inventors, and technology investors and journalists who have bewitched the modern world through inventions that speed up our daily lives without necessarily making them better."
In a recent interview with The Economist magazine, tech overlord-in-chief Elon Musk explained that "[i]f you have vast numbers of robots with vast amounts of digital intelligence, you have a sort of a quasi-infinite economy." In such an economy, "anyone can have anything they think of." He added that "money won't matter in 2036" because everyone will have a high income without working. This is a Star Trek version of the future, he explained.
Musk's views exemplify the tech industry fantasy of a push-button, voice-activated future world of abundance portrayed in the many Star Trek television series and movies. Any serious viewer of the series and movies knows that in the Star Trek future, poverty has been eliminated, people do not need money and in fact, everything you want or need is available to you.
Given that, a friend of mine once asked, "If you have a replicator and a holodeck, why would you need to go anywhere?" But Star Trek denizens do travel quite a bit in order "to seek out new life and new civilizations. To explore brave new worlds, to boldly go where no man has gone before.” (This was later updated in the Star Trek: The Next Generation series to say "where no one has gone before.")
Musk said that in his future of abundance, some people may choose to work, the way some people choose to have gardens in their back yards even though they can get all the food they need from the grocery store.
Now, here's the key element of this fantasized future: In 10 years AI will be running the world without humans. If that happens, the interviewer wonders, how will companies make money? How will they even exist? Musk has no answer.
And that leads to a second question: How will people get money to buy the things they need? Musk said that the government can just issue checks to people. This is an interesting view from a man who joined the Trump administration to cut government spending while heading its ill-defined and ill-fated Department of Government Efficiency. What Musk is proposing, it seems, is a form of AI-induced socialism. Many years ago he styled himself as a socialist. But, it's hard to take that claim seriously now given his publicly demonstrated political views and actions and the types of political candidates he is funding.
First, it's important to subject this fantasy of the future to what I'll call the "will" test invented by AI critic Ed Zitron. He asks AI boosters to explain what current AI models are capable of doing without using the word "will." Those advocates have little they can point to.
The tech overlord fantasy outlined above depends entirely on the word "will" with regard to AI. Nothing current AI models can do remotely fits this view of the future. Those models—called Large Language Models (LLMs)—produce so many errors that handing over mission-critical systems across the economy to AI as suggested by Musk would almost certainly create catastrophe rather than prosperity. The problem of so-called "hallucinations" is an inherent feature of the way LLMs operate and cannot be eliminated. And the kind of intelligence needed to run most mission-critical systems, that is human intelligence, simply isn't achievable with LLMs.
The idea that the current large tech companies that are controlling the development of AI are going to become extinct as worldwide prosperity spreads appears to be tale designed to lull the public into passivity. It makes sense that this narrative is okay with the industry given that the emerging literal killer apps using AI are surveillance and military weapons systems. Governments seem ready to pay the high cost of AI in order to obtain better surveillance of citizens and more effective weapons. That such systems will inevitably make many mistakes seems not very consequential to the governments that use them.
Businesses and consumers may be finding some benefit to using AI, especially for enhanced searching on the internet. But the true costs of such limited benefits is many, many times what they are currently paying. And given the way AI works, that is, requiring additional data center capacity for each new user, there are really no economies of scale that will bring down costs. Simply said, each new user costs the AI provider more money. AI as it is currently conceived is unlikely to be profitable except for niche users who find its output worth paying the true costs for.
The tech overloads have been singing the same tune for years: Trust us, sit back, and let us take you into an ever grander techno-uptopian future. New claims that AI will be the vector for this future are part of a propaganda campaign designed to keep you from having a say about AI and its uses. Fortunately, many communities are saying that they do not want to bear the costs and disruptions of new data centers associated with the AI build-out. That concern needs to be extended to the uses and abuses of AI itself.
Kurt Cobb is a freelance writer and communications consultant who writes frequently about energy and environment. His work has appeared in The Christian Science Monitor, Resilience, Common Dreams, Naked Capitalism, Le Monde Diplomatique, Oilprice.com, OilVoice, TalkMarkets, Investing.com, Business Insider and many other places. He is the author of an oil-themed novel entitled Prelude and has a widely followed blog called Resource Insights. He can be contacted at kurtcobb2001@yahoo.com.
No comments:
Post a Comment