I doubt that’s the case. My guess is we’ll hit asymptomatic returns from transformers, but price-to-train will fall at moore’s law.
So over time older models will be less valuable, but new models will only be slightly better. Frontier players, therefore, are in a losing business. They need to charge high margins to recoup their high training costs. But latecomers can simply train for a fraction of the cost.
Since performance is asymptomatic, eventually the first-mover advantage is entirely negligible and LLMs become simple commodity.
The only moat I can see is data, but distillation proves that this is easy to subvert.
There will probably be a window though where insiders get very wealthy by offloading onto retail investors, who will be left with the bag.
The administration doesn't decide spending. Congress does. There's no chance we get an anti-AI majority until a major AI catastrophe turns the public against it.
I just don’t see how they’ll be able to make a profit. Open models have the same performance on coding tasks now. The incentives are all wrong. Why pay more for a model that’s no better and also isn’t open? It’s nonsense
I wouldn't say the same but it's pretty close. At this point I'm convinced that they'll continue running the marketing machine and people due to FOMO will keep hopping onto whatever model anthropic releases.
If the frontier models reach a point of barely any noticeable improvements the trade off changes.
You do not need a perfect substitute if you are getting it for free...
People will factor in future expectations about the development of open source vs frontier models. Why do you think OAI and anthropic are pushing hard on marketing? its for this reason. They want to get contractual commitments that firms have to honour whilst open source closes the gap.
The person they were responding to said "Open models have the same performance on coding tasks now." AFAIK this is bullshit, but I'd love to be corrected if I'm wrong.
Open models, in actual practice, don't match up to even one or two generation prior models from Anthropic/OpenAI/Google. They've clearly been trained on the benchmarks. Entirely possible it was by mistake, but it's definitely happening.
lol bit of a stretch there, seeing as there are dozens of companies training LLMs.
As training software and infrastructure matures plenty more entrants will enter the market. It’s not like this is a particularly challenging research field, just very expensive at the moment.
Which LLM should I be using for programming work that isn’t released by OpenAI, DeepSeek, or Claude?
Which one outperforms this small handful of options within the AI oligopoly?
This statement you’re making is like saying “there are dozens of Android phone OEMs” when in reality Apple is gobbling up 80% of the profits, Samsung/Google are consuming another 10%, and everyone outside of China has their app installs gated by Google Play or Apple App Store.
Generally people in power will surround themselves with yes-men. It takes a good amount of humility and sincerity to look beyond this and deliberately choose people with a spine to listen to.
The most charitable interpretation is that most rich/powerful people are just as flawed as everyone else. Obviously, their power/wealth makes them less deserving of that charity ultimately.
> The most charitable interpretation is that most rich/powerful people are just as flawed as everyone else.
I can't believe that. They pulled themselves up by their bootstraps at their private schools and then had to claw and fight as a legacy admission to the school their parents attended. From there they lived hand to mouth destitute with barely a million dollar loan from their parents!
Then there was the existential crisis of meeting with their college roommates' parents and their own parents' bridge buddies to secure millions in loans. It was their flawless vision and skill that let them be at the right place and the right time. If they wouldn't have had the foresight to fall out of a lucky vagina we would all be worse off.
You see they're scrappy go getters that started from the absolute bottom. They're infallible supermen whose greatest assets are their humility and unerring genius.
> Generally people in power will surround themselves with yes-men.
It's probably a CEO thing too - you have some vision for the company so you're going to hire people that enable that vision, not people that will question your every move.
CEO is one of the least meritocratic jobs ever. It’s all just vibes, and the vibes are based off of what school you went to, who your parents know, where you grew up. Deep down they probably know this hence the insecurity. If it were a meritocracy they’d be toppled fast.
Can we agree that you are exaggerating? Not that you are totally wrong, but the flip side is, that ceos do need a different skillset. Workers who excel at the bureaucratic grind might not make the best leaders for lack of vision and empathy. Then again the concept of an empathic leader also seems to be foreign to many. It's hard to see anything with all the bullshit covering everything.
Eh, you don't hear about the ones who are well adjusted.
Usually nobody cares if you're a poor jerk. At least unless you do something phenomenal you don't get wide attention.
"New" rich people, especially those with power over other people, can develop plenty of complexes and insecurities that come out as weakness... like firing somebody for mocking them.
"Old" rich, generational wealth tend to develop a set of manners and habits where they don't get noticed or embarrass themselves quite so much by displaying such weakness.
You don't stay rich for a long time if you act like a fool.
I've been in positions to hire/fire before and never, ever would I fire someone over a kindergarten insult. How pathetic do you have to be to pull rank over "you rich jerk"
excellent case btw why you should never let these tech bros have power over your life, they're super charged angry little school boys with worse fantasies than a Soviet commissar
So over time older models will be less valuable, but new models will only be slightly better. Frontier players, therefore, are in a losing business. They need to charge high margins to recoup their high training costs. But latecomers can simply train for a fraction of the cost.
Since performance is asymptomatic, eventually the first-mover advantage is entirely negligible and LLMs become simple commodity.
The only moat I can see is data, but distillation proves that this is easy to subvert.
There will probably be a window though where insiders get very wealthy by offloading onto retail investors, who will be left with the bag.
reply