Yeah…and I suppose they all hauled ass out of there because their company was doing so incredibly well and they felt super optimistic about the near future?

Also, if you can’t make a chatbot BF/GF company profitable, that’s a pretty damning signal for the industry as a whole, certainly from an end-user perspective.

So all the rats abandoned ship when they saw that it was sinking. All the AI ships are sinking because there is no way to make it profitable. The operational costs are just too high. I imagine there are a lot of AI company CEOs watching this and deciding whether or not they too should be jumping ship now. Because it’s not a matter of if, it’s a matter of when.

No one in the AI industry is making money. They are all burning through venture capital just to stay afloat. They don’t have viable economic a plan, the pricing they would have to charge in order to make a profit makes the product unappealing to the vast majority of people. So either they continuously subsidize the product in which case they don’t make profit, or they charge what they actually need to charge and then they don’t have any customers.

It’s a no wind scenario created entirely by the fact that they are trying to put an in-development technology out into public before it’s really ready.

They’re doing a Tesla.

It means you:

• Pay off the investors by licensing the tech, so they get made whole
• Hire whichever employees you feel like taking from the team, typically the founders + their favourites
• Leave the rest of the company in place

You can do it as a founder as a type of exit where you get paid a whole lot of money.

The bad part is that every other employee at the company basically gets stiffed. The “social contract” of startups is that you forego a big tech-level compensation package in favour of stock options which may provide returns at the point of an exit. In reverse acquihires this social contract is directly violated.

With the added wrinkle that there are expensive things that may be necessary-ish, but when you exclusively use it for those things, the price jumps up to insanity.

Now that people are shunting easier requests to local models and sending the harder problems out, the hyper scalers have less income, so they have to charge more.

Because eventually the only real limitation will be hardware, that kind of AI will be a rarely-used commodity priced only slightly above what it would cost you to recreate it inside your business.

Not to mention the privacy concerns you’d have sending your expensive and necessary compute to the cloud.

Why is that? The electricity burned is about the same, no? Of course, the impact is more distributed.

I think the difference would be that local machines aren’t subsidized, so you’d see people get a lot more skittish after that first electric bill. Data centers are like railroad tracks from the railroad boom (if they’d made the rails out of licorice twizzlers.)

That doesn’t check out when considering model size. Unless that’s what you’re trying to say that self-hosted models force people to use smaller models. Just from an operations perspective a data center had economies of scale and the ability to batch up multiple independent user’s requests which should reduce the per token kWh along with the fact that your utilization will be higher vs somebody’s GPU at home would be idling.

That said I do believe there are negative environmental impacts from AI like always using giant frontier models for everything instead of smaller fine tuned models.

EDIT: I see a bunch of replies on the other comment, which all raise valid points, but that’s definitely not the way most readers are going to interpret the original post and moves the goal posts.

Didn’t say it was.

If you love someone, run them locally.