"What AI Won't Do"
by Bill Bonner
Youghal, Ireland - "AI...AI...ai yi yi! In the Financial Times there appeared a well-meaning piece of nonsense. “AI will do 80% of the economically viable work that humans do today, for 80% of all jobs, faster than most people believe,” writes Vinod Khosla, who claims to have spent decades brooding over the manner in which innovations upset the human animal. Brace yourselves for “mass underemployment,” he warns. And then he settles down to the true vocation of world-improvers everywhere - angling for a cut of the action. We require a “policy framework,” says he.
Why? Cannot the people work out for themselves how, and whether, to make use of the thing? Apparently not. They want busybodies of the Khosla stripe to arrange it all on their behalf, to rob the liquor store and then share out the gin. So, at least, Mr. Khosla believes.
More precisely, he would have the federal government seize a fat slice of the profits and the ownership of the AI industry - on the cheerful supposition that the casta politica will prove a wiser steward of the new marvel than the men who have poured billions of dollars and their whole careers into conjuring it up.
How likely is that? AI is run, at present, by some of the keenest minds on the planet - and even they have yet to show that it can turn a profit. Hand it over to the political hacks and you very nearly guarantee its ruin. Under Biden, the chief of the program might have been a diversity hire; under Trump, one of the family cronies. Either way, the odds of success grow slimmer by the hour.
The whole performance is the usual political smash-and-grab, dressed up in the time-honored uniform of problem-and-solution. Khosla holds that AI is a monstrous problem, since it will fling ordinary folk out of work while “juicing” the returns of the capitalists. We must, therefore, overhaul the tax code. And here the poor fellow betrays a truly remarkable poverty of imagination. Did the unemployed teamsters “juice” the earnings of the canal companies? Did the automobile makers fatten their books by putting the manure collectors out of business? Did the clothiers swell their returns by rendering the bobbin boys redundant? Of course not. No man “juices” his profits by throwing his customers out of work. Where would the juice come from?
So we begin by observing - modestly - that the dread now being peddled by these earnest soothsayers is misplaced. Only Washington and Las Vegas prosper by impoverishing their customers. The rest of capitalism, whatever its sins, makes the proletariat richer along with itself. Henry Ford doubled the wages on his assembly lines and made Detroit the best-paid city in the Republic. Today the highest wages are paid in Silicon Valley - again, very nearly double those of other towns. Even with AI hard at its labors - and here we speak with our fullest confidence - people will go on wanting things, and go on finding things to do to earn the money to buy them.
This is not to say the contraption arrives without its warning labels. No great innovation ever came to market without its risks. Every last one of them is, in the fullness of time, turned to the defrauding of men - or the killing of them. The first controlled flight took place at Kitty Hawk in 1903. Less than ten years on, Captain Carlo Piazza was dropping bombs on Turkish troops from his Blériot monoplane. America made the first successful test of an atom bomb in July of 1945; four weeks later it dropped one on the civilians of Japan.
The real danger of AI is that it will do something wicked - by design, or by accident. But that plane, dear reader, is already aloft - and at the controls sit the very people upon whom Mr. Khosla would bestow his “policy framework.” They are using AI to drop bombs and steer drones across the Middle East at this moment. And those same worthies have already ordered a ban on Claude 5, on grounds of “national security.” If anyone is going to bend the technology to scamming people, killing them, or robbing them blind, in short - it is going to be us.
Khosla’s economic menace is largely a fiction. His remedy is a fraud besides. And here, again, our confidence runs high - for the simple reason that both Bernie Sanders and Donald Trump have signed on to it. Sanders would nationalize the AI companies outright, that the common citizen might share in the spoils. And Trump throws in his two cents: ‘I’m going to have meetings with the top 12 or 15 executives very shortly, and we’re talking about giving back something to the public, and if we do that, the public will become very rich.’
Yes - they are counting the money already. Khosla and his fellow dreamers fancy that a blend of AI-linked taxes and an ownership stake in “public” hands will haul in “an estimated $150 - $200 billion within five years.” In ten years, it will be such a triumph as to furnish a “trillion-dollar annual kitty” to fritter away however we please.
Oh, how it carries us back. We recall a callow youth, ready to swallow very nearly anything. In 1969 we turned twenty-one, and Americans set foot on the moon. Surely it was only a matter of time before we kept a permanent colony up there - and pressed on to Mars. Fifty-seven years have since rolled by. There is still no permanent station on the moon. But Elon Musk pledges to plant a colony on Mars. And now the President himself assures us that Americans shall grow “very rich” - because the feds propose to poke their long noses into the AI business. How likely is that? Tune in tomorrow..."
No comments:
Post a Comment