Is it, though? We cannot predict technological advancement, and the times of ̶M̶u̶r̶p̶h̶y̶'̶s̶ ̶L̶a̶w̶ Moore's Law* for computational power are long gone. There is simply no guarantee that the costs will go down enough.
* thanks lucianbr!
The times for Murphy's Law for computational power are just beginning.
I don't expect bleeding-edge models to become any cheaper, but previous generation models can potentially be really cheap.
It would require multiple order of magnitude cost reductions to make that worthwhile. Maybe another few decades of Moore's law, if we have that left.
This was the Moviepass model of selling $10 bills for $9.
The Moviepass thing, I think if you were kinda gullible you could maybe buy into it eventually working on scale. This could never work on scale.
If you build a website that gives $100 for free to each one of your users, you’ll quickly have "demand" but that’s not "a matter of time until costs can be reduced".
Once the market consolidates, and users get used to the idea of using models that are "good enough" because frontier models are too expensive, there's no reason AI cannot be profitable.
AI is an inverse gold rush, the people who are getting rich off it are the people using it. The shovel-sellers are screwed.
The price of GPUs and the price of RAM to put in the servers.
Idk if Instagram would exist if they were spending hundreds of millions a day.
Paying near-infinity dollars for T-Shirts people want for $0 isn't a profitable business model.
Demand side price sensitivity impacts potential supply side margins.
And if nobody is willing to pay for it, it hardly matters how low you bring down cost, because it’s always a net negative.
Just look at the general state of the internet over the past two decades. Do you think it would work for Sora to insert ads into the slop?
My impression is that video is too complex to easily fit into an AI pipeline. Either you need something highly specific, like your own product’s UI. Or you need something personable and consistent, like someone talking into his camera.
The files are a pig to try and edit as well, making them beyond the generation and prompt costs expensive. At that point you might as well go and just film the ad.
Why spend the effort making a show for people on their phones? Will they even notice if it’s slop?
General business stuff like content or images has demand from across the economy. “Replace Hollywood” is kind of a niche thing.
$15m/day inference? How was that calculated? Forbes? Did they get it right? Is that a reasonable estimate? Still valid? How was revenue calculated?
IMO most of the votes had to come from some vote ring (35 pts in 35 minutes for a crap article, no way.)
Even the basics:
> Every second of video requires rendering hundreds of individual frames
Was probably only 24 or 30 frames, not multiple hundreds per second.
Sora was neither.
You forgot to strip the quotes from llm claude response.