LLMs are cheap
snellman.net
LLMs are cheap
1–10 of 319 posts
Re: LLMs are cheap
#2low to moderate quality digital text work is now almost free!
This is going to reshape large portions of our text based communication networks.
Re: LLMs are cheap
#3low to moderate quality digital text work is now almost free! This is going to reshape large portions of our text based communication networks.
There's of course also the issue that an increasing fraction of web content reading is being done by AI agents. I wonder what the Pareto front here is.
Re: LLMs are cheap
#4So far. Give it a few years when the core players have spent their way to market dominance and regulation kicks in and you'll see the price hikes investors have been promised behind closed doors.
Re: LLMs are cheap
#5You can't compare an API that is profitable (search) to an API that is likely a loss-leader to grab market share (hosted LLM cloud models).
Sure there might not be any analysis that proves that they subsidized, but you also don't have any evidence that they are profitable. All the data points we have today show that companies are spending an insane amount of capex on gaining AI dominance without the revenue to achieve profitability yet.
You're also comparing two products in very different spots in the maturity lifecycle. There's no way to justify losing money on a decade-old product that's likely declining in overall usage -- ask any MBA (as much as engineers don't like business perspectives).
(Also you can reasonably serve search queries off of CPUs with high rates of caching between queries. LLM inference essentially requires GPUs and is much harder to cache between users since any one token could make a huge difference in the output)
Re: LLMs are cheap
#6You can't compare an API that is profitable (search) to an API that is likely a loss-leader to grab market share (hosted LLM cloud models). Sure there might not be any analysis that proves that they subsidized, but you also don't have any evidence that they are profitable. All the data points we have today show that companies are spending an insane amount of capex on gaining AI dominance without the revenue to achiev…
For example, Perplexity has been fudging their accounting numbers to shift COGS to R&D to make their margin appear profitable:
https://thedeepdive.ca/did-perplexity-fudge-its-numbers/
Re: LLMs are cheap
#7> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low.
This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably harder than taking a user from $1 to $2, and the vast majority of those active users would drop as soon as you put a real price tag on it, no matter the actual number.
Re: LLMs are cheap
#8> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low. This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably ha…
There are more monetization ways than just a hard paying user. You can ask Google or Facebook. I dont think its super hard to get chatgpt to a. Profitable business. Its probably the most used service currently out there. And its use and effectiveness is immense.
Re: LLMs are cheap
#9> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low. This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably ha…
To make a billion dollars, I would simply sell a Coke to everyone in China. I have been giving away Coke in China and it is very popular, so I am sure this will work.
Re: LLMs are cheap
#10> OpenAI reportedly made a loss of $5B in 2024. They also reportedly have 500M MAUs. To reach break-even, they'd just need to monetize those free users for an average of $10/year, or $1/month. A $1 ARPU for a service like this would be pitifully low. This is a tangent to the rest of the article, but this "just" is doing more heavy lifting than Atlas holding up the skies. Taking a user from $0 to $1 is immeasurably ha…
Exactly, when the cost is free, I can ask it for whatever stupid thing I can think of.
The minute it starts costing me money, I have to make that decision: Is this worth the dollar?