COMP 2300 Winter 26 / Elle Hubert
We exist in a system where the main driving force is profit. These corporations are performing a balancing act between profit and optics.
Their negative effects are hard to measure, relatively easy to look at, and they are protected by the allure of the future.
The only reason that large companies don’t operate entirely off of slave labor is that it would reduce their bottom line. AI companies have a number of benefits that past companies have not. Their negative effects are hard to measure, relatively easy to look at, and they are protected by the allure of the future.
In ethical terms, AI fails on almost every metric. It is a tool that damages the environment, steals work from artists, and has the capacity to result in horrible tragedies, such as the suicide of a minor after engaging in sexual chats with a chatbot. Quite simply, AI has no moral ground to stand on as a technology.
Yet, because it profits, it is allowed all these things and more.
The point of this paper is not to argue against AI in an ethical sense. It is to argue against it in an economic one.
Nvidia
Nvidia, the self-proclaimed world leader in AI computing, currently makes up 16.2% of the GDP of the United States. It currently has a valuation in the trillions (yes, trillions with a T).
This is a massive amount of money, but the question has to be asked: where does it come from?
Compared to the rest of the companies on this list, Nvidia doesn’t have a store you can go to and buy things or products you see on the street: Nvidia makes computer chips, of which half go to only three customers. One of which is almost certainly…
OpenAI
And their new invention, a burning pile of money!
Nvidia has a lot of business with OpenAI. AI is the future, and Nvidia makes the highest quality materials. As long as OpenAI has money, Nvidia will sell them chips. Speaking of, let’s look at OpenAI’s money.
At first glance, this looks pretty good. 20 billion a year, directly correlating with spending? This is an investor’s dream. Except: this is a revenue graph, not a profit graph. Meaning, this graph doesn’t take into account all the money they spent.
20 billion a year? Try 143 billion NEGATIVE before making a single dollar of profit. And that’s not even close to what OpenAI has pledged to spend over the next decade. To put this in easier numbers to grasp: this is the same as spending 100 dollars over the next ten years for my company that has an annual revenue of $1.27, and has already spent $6. There is simply never a situation where this return on investment is matched.
This is why Nvidia has such a high valuation: any company would have it with the knowledge that you will receive a significant portion of a trillion dollars over the next decade.
Aside from this, OpenAI doesn’t seem to know a lot about their business model, or their customers (or, as it is beginning to seem, much of anything at all outside of making neural networks and generating hype). With the aforementioned ethical dilemmas (and many, many more…) surrounding AI, there is a solid amount of the global population determined to never use or spend money on it.
Assuming they had the entire world using it, with the current rates of people actually paying? They’d make 8.3 billion dollars per month… So only what, eleven years to start making profit?
Too Big to Fail
What happens when the bubble pops?
The term “Too big to fail” doesn’t mean that something can’t fail. It means that it cannot be allowed to fail. All possible action must be taken to prevent the bubble from popping. But it will.
At a certain point, AI will no longer be an unshakable certainty: it will be a financial novelty, on the level of NFTs. Investor opinion will turn, OpenAI will go under, and without it? Nvidia falls, not entirely, but still, leaving a loss of GDP not seen since the global pandemic.
Part of the reason that the 2008 financial crisis was so devastating was that there was no warning, there was nothing we could have noticed. Except there were. They were there, plain as day, and nobody did anything until it was too late. This is the same. We know this is happening, that it must happen. Better it happens within our control than without.
