> it's the running costs of these major AI services that are also astronomical
There's wildly different reports about whether the cost of just inference (not the training) is expensive or not...
Sam Altman has said “We’re profitable on inference. If we didn’t pay for training, we’d be a very profitable company.”
But a lot of folks are convinced that inference prices are currently being propped up by burning through investor capital?
I think if we look at open source model hosting then it's pretty convincing - Look at say https://openrouter.ai/z-ai/glm-4.7 . There's about 10 different random API providers that are competing on price and they'll serve GLM 4.7 tokens at around $1.50 - $2.50 per output Mtokens. (which by the way is a tenth of the cost of Opus 4.5)
I seriously doubt that all these random services that no one has ever heard of are also being propped up by investor capital. It seems more likely that $1.50 - $2.50 is the "near cost" price.
If that's the actual cost, and considering that the open source models like GLM are still pretty useful when used correctly, then it's pretty clear that AI is here to stay.
Lol anyway, the point is that even in a scenario where all the major models disappeared tomorrow (including OpenAI, Anthropic, etc), we would still keep using the existing open source models (GLM, Deepseek, Qwen) for a long long time.
There's no scenario where AI goes away completely.
I don't think the "major AI services go away completely" scenario is realistic at all when you look at those companies' revenue and customer demand, but that's a different debate I guess.
I've never worked at a company as large as Google but in my experience things can be a little more optimistic than the post. When earn enough trust with your leadership, such as at the staff/architect level, you'll be able to tell them they are wrong more often and they'll listen. It doesn't have to be a "$50,000 check" every time.
That leads to a very important question - Why doesn't leadership always trust their engineers? And there's a very important answer that isn't mentioned in the blog post - Sometimes the engineers are wrong.
Engineers are extremely good at finding flaws. But not so good at understanding the business perspective. Depending on the greater context there are times where it does make sense to move forward with a flawed idea.
So next time you hear an idea that sounds stupid, take a beat to understand more where the idea is coming from. If you get better at discerning the difference between ideas that are actually fine (despite their flaws), versus ideas that need to die, then you'll earn more trust with your org.
to some degree this is a "market correction" on the inherent value of these papers. There's way too many low-value papers that are being published purely for career advancement and CV padding reasons. Hard to get peer reviewers to care about those.
Yeah I don't know about that, the model providers like OpenAI, Anthropic, etc, literally sell intelligence as a product. And their business model is looking a lot more stable in the long term than all the startups built on top.
Simplest version of Kubernetes is zero Kubernetes. You can instead run your service using a process manager like PM2 or similar. I think even using Docker is overkill for a lot of small teams.
That's a good concept but I don't think Markdown is expressive enough for all the layouts & formatting that people typically want in PDFs. More likely that the source format would be something like HTML or SVG or .docx.
Restructured text has mostly 1:1 correspondence with Docbook. I use an XSLT transform to convert its XML schema into Docbook and PDF from there via XSL-FO.
Working on dev tools for MCP servers. As a building block I recently published a library to help write tests for MCPs - https://facetlayer.github.io/expect-mcp/
There's wildly different reports about whether the cost of just inference (not the training) is expensive or not...
Sam Altman has said “We’re profitable on inference. If we didn’t pay for training, we’d be a very profitable company.”
But a lot of folks are convinced that inference prices are currently being propped up by burning through investor capital?
I think if we look at open source model hosting then it's pretty convincing - Look at say https://openrouter.ai/z-ai/glm-4.7 . There's about 10 different random API providers that are competing on price and they'll serve GLM 4.7 tokens at around $1.50 - $2.50 per output Mtokens. (which by the way is a tenth of the cost of Opus 4.5)
I seriously doubt that all these random services that no one has ever heard of are also being propped up by investor capital. It seems more likely that $1.50 - $2.50 is the "near cost" price.
If that's the actual cost, and considering that the open source models like GLM are still pretty useful when used correctly, then it's pretty clear that AI is here to stay.
reply