Chinese open-source models are becoming the norm. You don’t need the latest model from ChatGPT (e.g., a Ferrari) to complete your task. A simple Honda will suffice (open-source models), and it uses a lot less energy since it runs on your home or company’s infrastructure.
Because of that, I predict that we’re going to see a lot of financial hardship among AI companies.
it uses a lot less energy since it runs on your home or company’s infrastructure.
A CPU-cycle requires energy no matter where it is. The open source models may be more efficient, or you may distribute the power usage over a greater geographical area. But you’re not using less energy because you’re running the model at home.
Because of that, I predict that we’re going to see a lot of financial hardship among AI companies.
I do like the idea that open source will kill tech giants.
You are correct. OpenAI etc. valuations are based on the difference between revenue and cost. The key part is that open source breaks the monopoly. It stops excessive rent being charged (now or in the future) to use Ai models.
Tech giants will still exist, but closed model advantages won’t.
If the valuation were based on the marginal cost OpenAI and Anthropic would be worth negative a trillion dollars…
Current marginal cost is certainly negative.
The expectation is that development costs will get cheaper (model design solidifies), Training costs will get cheaper (no need to retrain the whole model) and running costs will get cheaper (datacenter economies of scale).
All this is still true.
OpenAI, anthropic, Google etc. will generate excessive by being able to charge more than cost because their models are vastly superior.
Is likely to be false.
Datacenter economies of scale do not seem to be panning out, 'cause they’re buying into a RAM cartel who knows they’re desperate.
Coreweave’s reports say that the GPUs are 85% of the cost of a data center, and that they’ll last 6 usable years. This isn’t like building a facility like TSMC, datacenters feel more like COGS than capital to me, and those aren’t businesses I would want anything to do with even if I didn’t think they were all run by conmen who are rapists or racists or both.
I really want one or more of these AI companies to fail hard. I think their only strategy is literally to burn through billions hoping it’ll be another company that goes to the wall first and they’ll somehow reap the rewards by hanging on longer.
They’ve made it so they will all fail together. They’re bound. This way they can ensure a bailout, because if they don’t get a bailout, they will have a significant effect on the us economy on their way down.
Last week we received the CPU (Critical Patch Update) for Oracle products, which is a list of strongly recommended patches. They issue 4 per year (end of January, April, July and October). ALL patches were late.
You think/Are you saying they’re late because of the layoffs? Or is that the norm for Oracle?
It’s a trend. Years ago there were no patches finished after the main document. Then, it started appearing, to the point that they created a section warning about late patches. It was ok, because probably 99% of their clients must be on top of Linux x86-64. So, these patches were always available in time. This time, not even this patch was available. It was delivered a day later. Others, are scheduled for a week later, and sometimes, they reschedule.
I don’t have inside information. I only see the trend.
Years ago, patching a database was a big task, basically having to rebuild the dictionary. Then, they automated and changed stuff and things got super simple. Suddenly everything went downhill, to the point that last time I had to patch the patch.
I thought they were doing this on purpose to push us to their cloud service, but layoffs are a much simpler explanation. Occam’s Razor.
Couldn’t happen to a better group of companies. Let them all beg for bailouts and be denied.
and be denied
Maybe under a sane leadership, certainly not now. Trump will give them a bailout and a shipping container full of “deported” teenagers.
While I agree with the anti Oracle sentiment, that’s 21,000 employees that didn’t say they wanted AI BS. The suits and ceos that do this aren’t the ones that pay the price. It’s always the ordinary people like you and I.
