Mastodawn

Litrik De Roy May 11

Remember the early days of Uber and Lyft, when rides were dirt cheap because the companies were operating at a loss in order to capture the minds/wallets of the masses?

The rug pull in the AI/LLM world when the companies adjust pricing to actually make a profit is going to be spectacular. Especially when you consider the numbers of people / orgs that are addicted to or dependent on such technology.

⠠⠵ avuko May 10

@wdormann I have no mercy, but I will be forgiving.

digital luciferase May 10

@wdormann I think we're at the point where it's an opiate of the masses.

I gotta say I'm looking forward to seeing people go cold turkey, especially those who have the habit of taking LLM output and pretending it to be their own acumen at work. Or those who use ChatGPT to craft forum rebuttals and such - when this tide goes out, we'll truly see who brought their swim trunks and who never had them

*|FNAME|*May 10

@wdormann
This has been the dominant tech/startup model for decades. It’s incredible that people haven’t cottoned on this yet.

@witchescauldron

The Doctor May 10

@crispius @wdormann @witchescauldron And they scream at you for lying when you try to warn them. And again when they get screwed for not warning them.

random thoughts May 10

Companies and people don't care. They're looking at replacing people, especially at the junior level, but the junior simpler jobs are how people train up to be architects or experts. We're no longer giving them that chance.

The cost doesn't matter, as long as it's cheaper than a full time junior person being replaced.

@wdormann @witchescauldron

blackcoat May 10

@hittitezombie @crispius @wdormann @witchescauldron when I first started in this industry, decades ago now, I apologized for taking up so much of a senior engineer's time helping me solve a problem. His response has stuck with me ever since: "my most important deliverable to the company is more senior engineers".

The short sighted murdering of that pipeline in many modern orgs is *shocking* to me

Tero Hänninen May 10

@wdormann same for cloud computing

Francis ☑️May 10

@wdormann what if it took 5 or 10 years maybe until there were cost-reflective pricing for llms?

fedops 💙💛May 11

@wtfrank probably not doable considering they're burning not hundreds of millions but hundreds of *billions* now.

We're already seeing record unexpected spending levels, see the recent piece about Uber popping their entire R&D budget on tokens.
@wdormann

I am Jack's Found 404 May 10

Is uber still operating at a loss?

Last I heard they have yet to turn a profit because they're trying to undercut all the competition

Alex Turff May 10

@wdormann already starting to happen. See the recent GitHub copilot price increases.

Kierkethumbs up convincingly May 10

@wdormann except we’re not dependent on llm

Mother Bones 6d ago

@Kierkegaanks @wdormann Maybe not you and me 😌

Andyyyyyyyyyyyyyyyyyyyyyyyyyyy May 10

@wdormann next upgrade cycle everybody is going to want a laptop that can run local LLMs

@wdormann while I am critical about LLMs (but also use them) and it is clear that companies are currently trying to capture/create the market through losses, there is another scenario where LLM training/inference gets much cheaper through technological advances. But I am not sure of how probable this is.

John Abbe (aka Slow)May 10

@aaronkurz @wdormann Already happening. That's why Apple is suddenly out of Mac Minis and Mac Studios. Forward-thinking PC developers will be ready to pivot to that market when the current server boom busts. (You can be sure Apple is ramping up supply chains for relevant hardware in a massive way right now.)

@aaronkurz @wdormann A current trend among people really into AI is that they try to compensate for the inaccuracy by running the same text through multiple AI products, then asking one of them to combine and summarize the results. It seems plausible to me that resource waste will expand to compensate for technological advances, just like it has with RAM and disk space.
e.g. https://github.com/karpathy/llm-council

GitHub - karpathy/llm-council: LLM Council works together to answer your hardest questions

LLM Council works together to answer your hardest questions - karpathy/llm-council

GitHub

Vysogota May 10

@wdormann the plan is: try to keep your job until that moment. Then find another position at a company that fired all devs and got the rug pulled from under it. Demand the appropriate salary. It will be absolutely glorious!!

gunstick May 10

@wdormann they wait until all true developers have been eliminated, and then rise prices. Companies won't be able to hire back. Profit!

Administrator May 10

@wdormann Do you think companies that start installing LLMs directly on users' machines may have an edge in this war? Considering they're offloading the price of energy onto the end user?

https://www.thatprivacyguy.com/blog/chrome-silent-nano-install/

Google Chrome silently installs a 4 GB AI model on your device without consent. At a billion-device scale the climate costs are insane. — That Privacy Guy!

Google Chrome is downloading a 4 GB Gemini Nano model onto users' machines without consent, with no opt-in, no opt-out short of enterprise tooling, and an automatic re-download every time the user deletes it. The pattern is identical to the Anthropic Claude Desktop case I wrote about last month, but the scale is between two and three orders of magnitude larger. This article does the legal analysis and, for the first time, the environmental analysis. The numbers are not small.

That Privacy Guy!

Face Thumb May 10

@mdm @wdormann Yes. That will reduce the blood loss. You still have to train it, but you can train it once for all users, you can even distribute the training for the next model among your users too. The problem is efficiency, cloud providers *could* get efficiency through scale, use renewable energy, reuse cooling water. They typically dont, they do whatever is cheapest which is usually to freeload on the local town's resources.

@wdormann Yeah; I've been sandbagging on AI even though my org is all-in, partly because I really don't see the benefit, but _mostly_ because it was obvious what would happen as soon as AI was fully integrated into the workflow because _the exact same_ thing has happened with:
- Ride share services
- Streaming services
- Cloud services

When the heroin dealer comes around with "free samples", you don't proffer your veins, you stay away!
This is exactly the same business model.

@sleet01 @wdormann
That's chokepoint capitalism for you
https://en.wikipedia.org/wiki/Chokepoint_Capitalism

Chokepoint Capitalism - Wikipedia

@wdormann That's why I won't pay.

craignicol May 10

@wdormann @otfrom looking at the Anthropic and Microsoft/GitHub announcements, it's coming soon.

rvstaveren May 10

@wdormann haha, the same for the crack heroin that is called “AWS” ¯\_(ツ)_/¯

slackline :emacs: :orgmode:May 10

@wdormann Perhaps of interest https://www.wheresyoured.at/ais-economics-dont-make-sense/

AI's Economics Don't Make Sense

If you liked this piece, please subscribe to my premium newsletter. It’s $70 a year, or $7 a month, and in return you get a weekly newsletter that’s usually anywhere from 5,000 to 18,000 words, including vast, detailed analyses of NVIDIA, Anthropic and OpenAI’s finances,

Ed Zitron's Where's Your Ed At

Damien (TIG BUSINESS)May 10

@wdormann i'm expecting "too expensive for simple stuff, too terrible for complicated stuff" to be the eventual price point.

BenAveling 6d ago

Yup. It doesn’t matter that it doesn’t work well while it’s cheap. But once it’s not cheap, it has to work better than the alternatives. @aeduna @wdormann

@wdormann
I think the difference might be that Uber and Lyft provide actual services, even if they are leeches and screw their employees to death. Whereas AI doesn't provide anything of any real value anyway.

Quasit Most employers don’t want anything of value. They want to feel powerful, have a sense of control, and of being right all of the time.

LLMs provide all of that.

@kichae
Last I heard some companies had to start rehiring humans because AI couldn't do the work it was supposed to.

Very Human Robot May 10

@wdormann
I'm just using API pricing up front to make sure the economics still work out.
The $20/month plans will go away and the $200/month plans will be scaled back, probably.

@StompyRobot @wdormann
Eventually monthly plans will go away and everyone will have to pay per-token pricing
https://techcommunity.microsoft.com/blog/partnernews/github-copilot-is-moving-to-usage-based-billing/4515688

GitHub Copilot is moving to usage-based billing | Microsoft Community Hub

Instead of counting premium requests, every Copilot plan will include a monthly allotment of GitHub AI Credits, with the option for paid plans to...

TECHCOMMUNITY.MICROSOFT.COM

Morten Juhl-Johansen May 11

@wdormann So they will stop spending billions to earn millions? That sounds intuitively reasonable, but I understand that... building foundations... dynamic effects... next gen... mandatory in every toolbox... something

Mother Bones 6d ago

@wdormann I remember all the warnings at the time, and the heedless headlong rush. (Same with AirBnb). And then later the "gig economy" avalanche of problems due to that setup. An avalanche of problems that has never stopped.
So this should also be fascinating in the same horrible way.

@wdormann I'm pretty sure my employer already took my raise and paid it out in tokens instead of wages.

The tech I didn't want to begin with should have been paid for out of the executives' pockets, not the workers'.