stri8ted

Do they present any evidence to support these claims?

> You seem to be under the impression that somehow only US has the capability to advance this tech.

I did not say this, nor do I believe it.

> Aren’t these AI mega corporations constantly trying to scare us that if they stop pushing AI, China will get ahead? Which narrative should we believe?

Correct. If US paused, it will not be long before China exceeds US capabilities. I don't understand your question.

> Nah, they should be able to compete! To be honest, I find it hard to sympathize with trillion dollar ungodly rich corps.

You can't compete with free. No-one is asking for sympathy. I'm merely pointing out, people should not delude themselves that China is doing this for the goodness of mankind, and it will continue indefinitely.

Its not about bad/good. It's about self-serving interests. Currently, it serves China's interests to open source the models. If they had capabilities beyond those available in the US, I expect that would no longer be the case. Especially given the military/cyber implications. You don't expect them to open source their J-35 blueprints, so why would they behave differently here?

The reason this matters is, if along the way, they destroy all economic incentives for US labs, you may find yourself in a world, where you have no alternatives to closed Chinese models.

Look no further than what happened to mineral processing. Notice they are not open sourcing their mineral processing and magnet-manufacturing technologies.

I think you are underestimating the value of that last 10%. Returns to intelligence are non-linear. The 80/20 has been true for over a year now, and it has not impacted the labs revenue.

Except the revenues have moved in lock-step with valuations

FWIW, Hauku scores higher on AA Intelligence index, and is also cheaper per Intelligence index task. $0.21 vs DeepSeek $0.27

This model was likely trained months before deepseek released their paper.

I suspect the only reason those alternative providers have better up-time and more generous quotas, is because they don't have nearly the same amount of demand. Notice that Deepseek recently had to increase their pricing, once it gained it popularity.

> instead the queue of people in front of it will increase.

This can ultimately result in the system breaking. I don't think Anthropic engineers are that much worse than their peers, such that they are 10x more prone to causing outages due to bad deployments.

Reduced weekly limits, higher prices than competitors, paying well above spot price for space-x compute, all point to supply issues.

Demand > supply. It's impressive that customers have not migrated en masse to other providers, given the frequency of these outages. Perhaps switching costs are greater than some would believe. Or, qualitative differences between models continue to exist, despite matching on public benchmarks.

Translation, moderation, classification, guardrails, etc..

This is brutal, especially for small transactions. Note each failed attempt is another screening fee. Unfortunately, there is no good competition.

Demand > supply

Same

By what measure is there an overbuild? Every metric I look at, shows inference unable to satisfy current demands.

20k is small potatoes for the marketing impact.

They are doing both. Distilling Mythos down to affordable models, so they can continue to fund the business. And training Mythos level models at the high-end, to expand the frontier.

You seem to assume training on pelican would not result in improved performance on other similar tasks. Why?

Still has a 1.78T market-cap. It will be OK.

Its actually useful, when you are launching into a long monologue and want periodic acknowledgement that its "listening".

Way too expensive. Google vision OCR (which they failed to compare against), is $1.50 per 1k pages. Vs $4 from Mistral.

What metrics are you using to classify it as a downfall?

Is there something in the post that you find implausible or don't believe to be true?

I doubt this is representative of real world usage. There is a difference between a few turns on a web chatbot, vs many-turn cli usage on a real project.

Entire segments of the podcast sphere are making their money talking about these so-called unspeakable subjects. Why don't you share what you really think.

Those are the same thing

48 GB is not consumer hardware. But fundamentally, there are economies of scale due to batching, power distribution, better utilization etc.., that means data center tokens will be cheaper. Also, as the cost of training (frontier) models increases, it's not clear the Chinese companies will continue open sourcing them. Notice for example, that Qwen-Max is not open source.

Do you have any evidence to support this view?

Price Input: $2.50 / 1M tokens Cached input: $0.25 / 1M tokens Output: $15.00 / 1M tokens

https://openai.com/api/pricing/