Hacker News Reader

Comment by xscott | Hacker News Reader

I'm not imaginative enough to see more use cases and who will pay. You can burn a lot of tokens and dollars on math proofs. Same for white hats finding vulnerabilities in all the software. Generating video.

Personal assistants dealing with your email, filling out paperwork, and booking reservations don't use a lot of tokens.

If every software developer paid $200/month, every academic paid what they pay for Matlab or Mathematica licenses, every YouTuber paid for whatever video processing, and every normal person paid a subscription comparable to their TV subscriptions and cell phone plans - does that justify the valuations?

The government pays for surveillance? Companies pay for customer service replacements and target advertisements?

It seems like there are a few small categories that use a lot of tokens, and a large category that doesn't need many at all.

View on HN

> If every software developer paid $200/month, every academic paid what they pay for Matlab or Mathematica licenses, every YouTuber paid for whatever video processing, and every normal person paid a subscription comparable to their TV subscriptions and cell phone plans - does that justify the valuations?

> The government pays for surveillance? Companies pay for customer service replacements and target advertisements?

Yes definitely. All together that's probably within striking distance of a trillion dollars in annual revenue. Just TV subscriptions + cell phone plans in the U.S. alone is about 400 billion (leaving aside any other country).

Most large, mature tech companies are at 5-10x revenue to valuation ratios (and 10-20x EBIDTA to valuation ratios).

Of course whether you get there is a separate question, but any company that fulfilled what you're asking would pretty easily justify multi-trillion dollar valuations.

You make it sound reasonable then.

That $400B for video and phones though. That's like a $1k average for every adult and child in the country. It's twice what I spend, and I'm not on a family plan.

CTIA gives $235 billion for consumer mobile data revenue (CTIA only has a detailed summary for its 2025 report, which gives $224 billion for 2024, https://api.ctia.org/wp-content/uploads/2025/08/2025-CTIA-Su..., but second hand reports give $235 billion for 2025, e.g. https://www.lightreading.com/wireless/inflation-mostly-skipp...).

Video is a bit trickier to nail down since there isn't a single authoritative group putting out numbers, but most whitepapers converge around $150-200 billion (see e.g. https://www.prnewswire.com/news-releases/total-subscription-...).

I probably should've worded my reply better (eternal problem of mine). I wasn't doubting you so much as being amazed people spend so much.

Thank you for the detailed links though!

If every software developer paid $200/month

That's significantly underestimating the potential. Companies typically spend $200k/year for developers (salary, benefits, taxes, equipment, office space...). If AI can make them even 25% more productive, that would justify $50k/year in tokens.

I more than agree on the value, but not on the price. There's currently a competitive market, including free models that are pretty impressive if not "frontier".

I don't think LLM inference is ever going to stop being a competitive market, though. Unless regulation or other factors make the existing players form a cartel.

I think it's pretty safe to say that, unless the government steps in to prevent it, they will form a cartel or some sort of oligopoly. The incentives for them to do so are very high.

We're also one high profile cyber attack, possibly false flag, from the government granting that cartel.

Regulation wouldn't solve the problem, but many frightened people would be in favor of it anyway.

When the internet was just getting started, I imagine it would have sounded ludicrous to suggest that soon, practically every single software application would have some networking component.

Similarly, I think it's a good bet that in the relatively near future, pretty much all consumer software, down to piddling little phone games, will be built with the presumption of some LLM/inference layer.

One very simple reason for this is that AI could dramatically change people's expectations for UI/UX across the board. I can see a future where the expectation is to have some natural language interface at minimum, or have UI driven a great deal by inference under the hood.

In which case, applications without these features become about as appealing as command line tools to the general public.

> it would have sounded ludicrous to suggest that soon, practically every single software application would have some networking component

Still does.

If they do for material sciences what they did for math proofs, the world is going to get weirder!

Yeah, we need a lot more kinds compilers like LEAN. Give the accountants, physicists, chemists, biologists, and so on the tools where they can verify the results.

if we were able to make a LEAN for hard science we would all be living to 200 by now

Why so short? :-)

The labs probably foresee things looking more like OpenAI solving Navier-Stokes: a company identifies a strategic opportunity locked behind an extremely difficult problem, throws a huge amount of money and compute at it, solves the problem faster than their competition, snatches the opportunity.

Current companies are paying way more than 200$ per month per developers. In startups, SMEs and big tech, you can multiply it by x10

It seems like there are a few small categories that use a lot of tokens, and a large category that doesn't need many at all.

It won't be about tokens in the long run, and I think the labs know that. It will be about long-term memory and user relationships, and who gets to own that data.

The disruption and upheaval has barely even started, IMHO.

You mean they hope to be the next Facebook/Google that knows everything about you and call sell that data to advertisers? Yay :-|

Reply on news.ycombinator.com