The Cost of AI

There are plenty of news stories and analyses about the cost of AI, but not many discuss the opportunity cost of AI.

Last year the hyperscalers spent $450 billion on AI infrastructure, this year it is set to be $900 billion. That is a genuinely ridiculous sum to spend on anything, let alone a technology which appears not to be profitable to run, nor especially beneficial in terms of productivity. I’m not the first one to point this out, nor will I be the last; rather this post is an effort to put the abstract into concrete terms. It is easy to see that $450 billion is larger than $900 billion, but the comparison is meaningless until you see what a mere fraction of $450 billion can buy.

AI data centres, chips, and the other ingredients powering the generative AI boom are often described as infrastructure, so to begin let’s compare AI infrastructure with the less sexy infrastructure megaprojects of the past. The Suez canal cost 433 million francs and was inaugurated in 1869. According to a site for converting historical prices, that amounts to 4 724 620 831 American dollars in 2015, which is equivalent to $6 656 214 274 today. Or to state it in plain English, last year’s AI infrastructure buildout was about 67.6 times more expensive than the building of the Suez canal adjusted for inflation (the figure for the Panama canal is $12 521 925, or about 36 times more expensive).

Another point of comparison is the US federal budget. $450 billion is the cost of veterans’ benefits and the entire justice system, and $900 billion is a full $100 billion more than the US federal state spent on national defence this year. Pablo Escobar famously offered to pay off the national debt of Colombia in exchange for a pardon when his drug empire was at its zenith, and the US tech companies could have made Colombia’s creditors whole for a “measly” $330 billion, according to CEIC Data. The Kiel Institut’s data on the military aid to Ukraine from all states which sent aid since Russia invaded similarly landed on a total of $329.5 billion.

I’ll leave it up to readers to decide whether server halls are a better use of funds than supporting Ukraine, building canals, and funding governments – the point is that this is the scale of money we are talking about when it comes to AI. As is well documented on the blog, I think the AI mania will inevitably end in tears, and this is to me another example in an embarrassment of riches showcasing the disconnectedness of Silicon Valley elites.

If you liked this post you can read a previous post about the war in Ukraine here or the rest of my writings here. I also have a section for longer reads I call essays here, I particularly recommend my series called The Bird & The Technoking exploring Elon Musk's takeover of Twitter, and its political and cultural implications. It'd mean a lot to me if you recommended the blog to a friend or coworker. Come back next Monday for a new post!

---

I've always been interested in politics, economics, and the interplay between. The blog is a place for me to explore different ideas and concepts relating to economics or politics, be that national or international. The goal for the blog is to make you think; to provide new perspectives.

Written by Karl Johansson

---

Sources:

Cover photo by Olia Danilevich from Pexels, edited by Karl Johansson