pfannkuchen
Joined 8 years ago
Joined 8 years ago
Like are the LLMs so expensive to use that they are just converting employee spend to LLM spend? If so, that’s still odd, because the LLM capabilities just represent a massive new investment opportunity, and I think the same game theory structure would still hold unless the companies had minimal cash reserve and minimal lower ROI investment areas to cancel first (and unless something changed recently many companies do have cash reserves and also I haven’t noticed experimental projects getting canceled outright to focus on ramping up the main project into areas previously unreachable).
On the other hand, money got a lot more expensive around the time the layoffs started, and that would force everyone’s hand at once to reduce expenses, and it looks a lot better for various parties if the reason is OMG AI and not a roller coaster monetary policy.
Like if we jack up mortgage rates to 100% housing prices would drop massively right? Okay at some point mortgages weren’t widespread and now they are, therefore a large percent of the increase in housing costs comes from mortgages enabling the bidding up of housing and land by pulling future earnings into the present. The effect on land is zero sum, and while the effect on the structure is positive sum to a degree (it’s a big contributor to housing getting bigger than it used to be and bigger than it needs to be), the overall effect is negative and so we would be vastly better off banning the practice. Of course that will never happen, but I find it curious that it isn’t even discussed.
I keep checking back on LLMs every couple months and while they are now great for figuring out bugs in a system I don’t know much about (say in my use of a framework I’ve never used before), it’s been super helpful. But whenever I have it write a lot of code, even in a not huge project, it mostly works but it’s like a gum and popsicle stick machine gun and I’m scared of what large org code bases will look like in a few years.