AI Is Burning the Planet for No Good Reason
First, sorry for the lack of newsletter updates. I had to go into the emergency room for dehydration on Monday. Long story, but still not dead! Even if eating is still difficult. I swear, I am going to have a newsletter section called “What Tried to Kill KC Now?” just to document all the ways my body is trying to kill me.
With that bit of whining out of the way, I want to highlight a couple of excellent journalists and discuss how their different stories collectively condemn the tech and financial leadership that has led to this terrible outcome.
First is this article by Brian Merchant, author of the book Blood in the Machine. It is an excellent history of the Luddites and their echos in today’s economy that I highly recommend. His newsletter is just as well-researched and thoughtful. The post in question brings together several threads about how imitative AI is serving to revitalize the fossil fuel industry, just at a point when it looked as if we had reached peak fossil fuel usage:
Got that? We were, according to this report, on track for fossil fuel use to top out—until AI came along. AI is a notoriously large energy suck—an AI-generated answer needs about 10 times as much electricity as a Google search. Now analysts and agencies are quietly revising their decarbonization goals downward, gas and coal plants that were slated for retirement are being kept online, and now utilities are building more gas plants in the first half of 2024 than were built in all of 2020 combined.
…
So the AI companies are exacerbating the climate crisis in at least three ways:
Selling tools to help oil companies AI tools to accelerate fossil fuel extraction.
Running data centers that require ~10x the amount of electricity as Google
Giving fossil fuel companies a reason—or an excuse—to build more power plants.
Now, life is about trade-offs. One could argue that the energy requirements are temporary, and they the benefit that imitative AI provides outweighs the temporary slowdown in dealing with the climate crisis. It is not a convincing argument, but it could be a viable argument is imitative AI had any significant value. But it does not.
Which brings me to Ed Zitron’s piece about just how little imitative AI is producing, either for the companies that create the systems or the companies that use the systems. I encourage you to read the entire piece — it really is a well and thorough documentation of just much of a bubble imitative AI appears to be — but here is a decent summation of the argument:
In short, OpenAI is burning money, will only burn more money, and to continue burning more money it will have to raise money from investors that are signing a document that says "we may never make a profit."
As I've written about previously, OpenAI's other problem is that generative AI (and by extension the model GPT and the product ChatGPT) doesn't solve the complex problems that would justify its massive costs. It has massive, intractable problems as a result of these models being probabilistic, meaning that they don't know anything, they're just generating an answer (or an image, or a translation, or a summary) based on training data, something that model developers are running out of at an incredible pace.
…
And even if they did, it isn't clear whether generative AI actually provides much business value at all. The Information reported last week that customers of Microsoft's 365 suite (which includes things like Word, Excel, PowerPoint and Outlook, and more importantly a number of business-focused software packages which in turn feed into consultancy services from Microsoft) are barely adopting its AI-powered "Copilot" products, with somewhere between 0.1% and 1% of its 440 million seats (that’s $30 to $50 per person) paying for the features. One firm testing the AI features is quoted as saying that "most people don't find it that valuable right now," and others saying that "many businesses haven't seen [breakthroughs] in productivity and other benefits" and they're "not sure when they will."
This is not necessarily new information — I have talked about the lack of utility of imitative AI and professional market watchers are also warning about a bubble in the space but Ed does a spectacular job tying all the ends into a neat, if depressing, bow.
Imitative AI does nothing to justify the immense cost to the rest of society that it imposes. Now, every time I say something like that, I get people coming out of the weeds to insist that imitative AI can too be helpful. And I am not arguing otherwise. At the edges, there are things it can do decently enough at today’s prices that some people in some places may find it somewhat useful. But as noted above, today’s prices are a giveaway by the AI companies hoping to lock in customers, meaning they must go up at some point. And imitative AI is not proving itself to the majority of its users to be the kind of tool that justifies large investments. It simply isn’t producing — not at the level necessary to be profitable and certainly not at the level to justify the damage done to the climate.
All of this is a massive failure of society. If we did not pretend that shareholders where the only stakeholders in companies, we might not have MBAs are Google, Microsoft, etc. running the companies with the intention not of making great products but of buying back shares. If we actually imposed the cost of the climate damage on the companies producing that damage, then the cost would likely deter such wasteful extravagance. If we enforced anti-trust laws and kept companies small enough to democratically managed and encouraged real competition in the tech sector, then the bad ideas of a handful of executives and VCs would not necessarily dominate the sector.
But we don’t do any of that. And as a result, we are reducing the progress we have made in saving the only planet we have to live upon in order to prop up an industry that has no commensurate value. It is just auto complete, deep fakes, and misinformation all the way down. Our society is simply broken, at least when it comes to handling business’s effects on the wider public good. Maybe, finally, we can see that and do something about it.
Because I really don’t want the epitaph of the human race to be “They burned their planet for the sake of Fancy Clippy.”
Also, you should subscribe to Brian Merchant and Ed Zitron. They are excellent journalists who deserve support. And buy Merchant’s book — it is excellent.