One thing to remember is that the $200/month subscription is heavily subsidized. It is more to promote use, especially to corporate users that pay for the API token use.
> Still I miss when Google was a good information retrieval system
I think a large part of the blame is not on Google but on the websites themselves. The Internet has been enshittified by a gargantuan amount of spam sites and content mills created just to generate clicks and boost SEO.
At least AI offers a way to filter out the noise at the cost of relying on how it was trained and what the creators thought is good data.
They constantly reward websites that are on a hamster-wheel of chasing the latest SEO trends, while penalizing websites that have actual information for not jumping through the same hoops.
A company I know operated TWENTY THOUSAND blogspam blogs out of a single server/IP. Google knew all along that this was happening (the companies had strategic partnerships) and never did anything about it.
The last thing they did anything significant was when, Panda in 2011?
At this point it's clear they're a monopoly and only care about websites who cater to their whims + making money. Search be damned.
Google knows exactly what it is doing. The downranked .edu domains which always ranked high in 2005. They want to feed people with rage bait and SEO websites, since the persons who read that garbage are the only ones who react to advertisements.
That one is 100% on Google. They intentionally enshittified their own search so that you spend more time in it. And they made it harder for genuine sites to be found unless they play the SEO game.
Inference is cheap. I bet the financials of these Chinese companies are much saner looking than any of the big US AI companies which are bloated by investors.
DeepSeek is very likely selling tokens at a loss. There're many cloud providers that provide you with DeepSeek V4 Pro via API, and those services at least twice as expensive as DeepSeek itself.
I see no evidence anywhere that "inference is cheap". To my knowledge this is a myth being spread to pretend ChatGPT or Claude will one day make any economic sense.
DeepSeek likely operates at a loss. How big the loss is anyone's guess.
Meanwhile I am happy using their model. It is really good, to a point I forget I am not using Codex or Claude.
DeepSeek hasn't raised enough money to be actively selling tokens at a loss. They have a small team, extremely low overhead relative to other labs, operate in a place with the essentially the cheapest commercial electricity rates in the world, and their architecture lends itself very well to cheap inference.
If you think heavily subsidizing AI models isn’t financially viable, I have some bad news for you about US AI companies.
Deepseek has made some incredible advancements in model efficiency, and more importantly actually publishes those advancements so everyone can benefit from them.
Maybe not. I don't see how US inference providers can compete anyway with commoditized models. Costs are out of control here and the infrastructure is way worse.
They might be thinking, we already have the servers and the GPUs sitting there anyway so why not make full use of it? They're not even close to being at a mature state where they start to monetize.
For sure. But also they’re building an electrostate with 100% electricity redundancy and dirt cheap electricity. They might actually be able to sustain this.
US suppliers are fine and won't go bankrupt, they can just focus on serving bigger "Pro" class models from their large datacenters. In fact cheap AI makes the bigger and smarter models more useful because it's smart enough to draft a clear question to the model, which helps minimize wasted tokens.
> US suppliers are fine and won't go bankrupt, they can just focus on serving...
For a while, US automakers thought the same of Japanese, then Korean car manufacturers, and Musk laughed at Chinese EV makers in an interview >12 years ago. People learn and get better at making things until they catch up with the frontier.
Chinese EV makers have a few interesting technologies especially wrt. batteries but they're still very far from catching up to the frontier in a general sense. From that narrow POV Musk was absolutely correct.
What the hell are you talking about? They have batteries that charge 0-80% in 5 minutes even at -30F. More full featured EVs at half the price with similar acceleration rates and higher top speeds. Total ranges are comparable or better. What is this frontier you speak of? I think the only thing US companies are far ahead on is self driving.
US providers are burning VC money because they have been selling the idea of total world domination. Even the government has bought into that. Now suddenly they are not longer dominating the field and even need uncle Sam to protect them from foreign competitors.
They can still dominate wrt. the biggest and smartest models. DeepSeek does effectively nothing to change that. Of course these big models will be served at a very steep price in order to fully and completely recoup the investment, but there's no reason why that couldn't work if they really are smart enough and if the market value of smarts follows any kind of scaling law.
But who'll pay for those ads? Why would I pay Google if it just plops out some LLM answer with maybe my site as a source - which 90%+ of people will ignore as they don't care about it.
Is this any different than current search ads (as far as people ignoring them)?
I think this is a lot better for Google. It's always at the top of the page, harder to block with an adblocker, possibly more trusted (we'll see...)
You pay per impression or per click, just like in all the rest of their ads. Except these have a higher CPC since you're the only featured brand to go along with paragraphs of text about why your shoes are the best.
Yep. The thing is people (maybe because of our limited scope) just focus on the depth and not the breadth. Because this is a general purpose model - it also has PhD+ knowledge in Physics, Biology, History, etc.
I think we still don't really comprehend how much can be achieved by a single "mind" that has internalized so much knowledge from so many areas.
I've had the $20 Gemini plan to use when my local setup runs into tougher problems and the throttling today has been bonkers. I canceled my subscription and will look into upgrading my local setup.
Most Chinese companies will avoid Nvidia Gpu and as much american tech they can now when it comes to serving AI as now they know it can be stopped any time by the US or maybe even their own government so the risk premium is too high. They might still use Nvidia to build the models but not for running them and serving to customers
Up to the H200 iirc, but they haven't made a purchase yet afaik. The experts in such things believe if they do make a purchase, it will be a token one. Xi is pushing hard for indigenous production, not becoming "hooked" to American Ai chips like some (not so bright people) think we can cause to happen.
Doesn't need to be the Chinese. It can be anyone without stratospheric Nvidia margins. The Gold Rush phase of AI economy (aka "the bubble") is beginning to slow down and the Optimization phase is just beginning to ramp up (we see this with massive bumps to token cost and token burn rate of pretty much all frontier models, plus the general pivot away from your typical individual chat end-users to businesses and employees of said businesses) and there will come a time when "nvidia has the best software stack" will not mean much for the big players. Organically, I think it already kinda does, it's just masked with the inertia of massive circular deals and Nvidia selling its services to itself (entities it backs/invests in).
The availability of open models with such capabilities are based on the goodwill of the Chinese. And that might end eventually, especially that the matter is one decision of Xi and the party.
True but I’d argue they are good enough now to do 10 years of continued workflow automation. It’s like the internal combustion engine or personal computer, at some point they were good enough for broad categories of work. I think that’s where the current models are.
This is correct but people with too big of an ego or affected too much by Dunning-Kruger) will try to say otherwise even when presented with ample evidence. Instead of a valid response you'll get "skill issue" from people that produce segfaulting code on a regular basis.
reply