LLM are an embodiment of the Pareto principle. Turns out that if you can get an 80% solution in 1% of the time no one gives a shit about the remaining 20%. I agree that’s terrifying. The existential AI risk crowd is afraid we’ll produce gods to destroy us. The reality is we’ve instead exposed a major weakness in our culture where we’ve trained ourselves to care nothing about quality but instead to maximize consumption.
This isn’t news really. Content farms already existed. Amusing Ourselves to Death was written in 1985. Critiques of the culture exist way before that. But the reality of seeing the end game of such a culture laid bare in the waste of the data center buildout is shocking and repulsive.
The data center buildout feels obscene when framed this way. Not because computation is evil, but because we're burning planetary-scale resources to accelerate a culture that already struggles to articulate why quality matters at all
There isn't nearly enough AI demand to make all of these projects turn a profit.
Unless you successfully make ai subscription an obligatory cost to pay for every employee in an org. Which they seem to be doing quite successfully considering enterprise sales activity.
Successfull b2b salesmanship does not require a working and useful product. It just requires you to get these meetings with purchasers and show that other comparable orgs are buying your product. For example, at my last job I wasn’t sure if we were a microsoft shop or a google shop because we bought both of their products that did the same thing. Because that is just what you do when you are an org with 5 figure employee counts, you spend budget.
Very well put, one of the more compelling insights I've seen about this whole situation. I feel like it gets at something I've been trying to say but couldn't find the right words for yet.
Quality. Matters.
It always has, and it always will. If you're telling yourself otherwise, you are part of a doomed way of thinking and will eventually be outcompeted by those who understand the implications of thinking further ahead. [ETA: Unfortunately, 'eventually' in this context could be an impossibly long time, or never, because people are irrational animals who too often prioritize our current feelings over everything else.]
If quality matters then why is everything crap? Price has a quality of its own.
As I said in my edit, "because people are irrational animals who too often prioritize our current feelings over everything else."
Marginal improvements in quality which result in a marginal increase of cost/price often provide much better overall returns than just using a series of cheap substitutes that fail quickly. In some areas, this doesn't work, but I think shortsightedness is blocking truly better solutions in a great many cases. Particularly when true costs are being externalized.