By Audrey Elwood, Chief Columnist
The AI world is betting on itself to fail.
This is clear after a high-level Anthropic researcher sounded the alarm that AI’s current trajectory is “genuinely frightening” and that AI companies are “are racing straight to self-improving superintelligence and gambling with our lives.”
This is not a new concern, but this incident has sparked a new discussion around the fast progression of AI, leading major AI CEOs to issue statements advocating regulation of AI.
Anthropic’s CEO has issued statements saying AI will have the power to kill the world’s population by 2030, just four years away. While there are talks and calls for regulation from CEOs, nobody wants to slow down. If you are not leading the charge, you are left behind and that is not great for business.
Pascal’s wager is a famous thought experiment that is often cited as a defense of Christianity. The argument is as follows: if you do believe in God, if heaven is real, you benefit, and thus you do not lose anything by believing in him; if you do not believe it, you could suffer eternal damnation or have nothing happen.
So, under this theory, any “rational” person would believe in God; you win everything if God is real and you lose nothing if you do not – so why wouldn’t you believe in God?
AI companies are making the same bet; they know we are marching to doomsday.
Chief Columnist Audrey Elwood finds that artificial intelligence leaders are making a metaphysical gamble in the pursuit of unknowable and irreversible change
The massive acceleration of AI would make any regulation moot by the time it could be passed. We are on a runaway train, and if you are not on the train, you will get left behind.
If AI kills us, we will all lose. If AI doesn’t kill us and you were on the train, you will be filthy rich.
If you are not the one contributing, someone else will; and you will be left behind.
A prominent theory in Silicon Valley right now is of the permanent AI underclass, in which people who did not adopt AI, or are in AI-replaceable jobs, will be pushed into extreme, inescapable poverty.
These are the figures tech bros shout out when they are telling you to become a plumber instead of an investment banker.
With a recession of looming, gas prices creeping up and generally poor economic sentiment, this thought gets scarier and scarier. This is especially true for the United States, where the American dream ideology still looms.
If you cannot pull yourself up by your bootstraps, if you are trapped in a permanent inescapable struggle, there is no point in being American.
We can see parallels between the AI crisis and the climate crisis. We are now at the point of no return with the environment, and a large reason there was little effort made to solve this disaster was that nobody wanted to be left behind.
The thought of making sacrifices to help the climate, then getting left behind; yet losing status and economic development outweighed the threat of acceleration toward an unknowable future.
But it is not the developed countries that lose; it is the countries with fewer resources and, because of the lack of resources, has worse recovery chances.
The same goes with AI: the companies that are wagering with humanity have bunkers and contingency plans if things go wrong.
If AI does create a permanent underclass, they will not be the ones cooking squirrels over a barrel. They will be kings. AI has been integrated into our lives so much that it is unavoidable unless you go off the grid. There is no stopping it.
There is potential in creating safeguards, but as we have seen time and time again, there is no incentive for these safeguards’ unless a wide-sweeping change happens rapidly, which is effectively impossible.
Pascal’s wager on AI is, unfortunately, being lived every day: as few people gamble with our future, being realistic, you have to hope you can hop on the train or be steamrolled. AI is part of life; the only choice is to get in or get left behind when the AI rapture happens.