> $120K isn't going to cover the fully loaded costs of an SRE who can set up and run that.
> Hiring 1 person to run the infrastructure means that 1 person is on-call 24/7 forever.
> If there's an issue with the server while they're sick or on vacation, you just stop and wait.
Very much depends on what you're doing, of course, but "you just stop and wait" for sickness/vacation sometimes is actually good enough uptime -- especially if it keeps costs down. I've had that role before... That said, it's usually better to have two or three people who know the systems though (even if they're not full time dedicated to them) to reduce the bus factor.
So the entire business was happy to go offline for 2/3 weeks whenever their infra person fancied going off on their summer holiday?
By doing this, you're guaranteeing a bus factor of below 1. I can't think of any business that wouldn't see that as being a completely unacceptable risk.
I never understand the drive to stay away from cloud services for small scale operations. It’s not your money that’s being spent on the cloud, but it is your free time being asked to be on call when you encourage your company to self-host!
Bus factor 1 is rarely enough for "entire business". But if the GPUs are for training models, and their users are the data scientists that are also on holiday around the same times - that might indeed be good enough policy.
Ouch, that is indeed a risk one must be wary of. Can be a "works for the company but sucks for employees". Which can also drain the company of skilled people, a poor trade in most cases.
> The one place where OpenAI does have a clear lead today is in the user base: it has 8-900m users.
There is no way that number is an accurate reflection of the number of actual human users of their service. I could believe they have 8-900m bot/fraud accounts in their databases, maybe, but not real users.
I suspect I am one of those 1bn users by their metric. I have an account, and I sometimes query it. I also query Claude and Gemini. I have zero loyalty, if I run out of tokens on one, I will just pick up the conversation on another provider. Perhaps I am using them wrong, but the amount of babysitting I have to do anyway, I don't find it that tedious to stay on the same topic during a swap.
There's no way I would spend $200 a month on any of them, not even $20 considering how few 'tokens' you get. I can see how these tools would be useful to my workflow, but I cannot use them as they are priced 100x too high for me to be reliable.
I have a feeling that would be true for the vast majority of these AI tool users. I really am not sure how these companies are supposed to become profitable. But SV is a bit insane that way.
i used to use them that way until I got a IDE with AI and is way better for my development than copying and pasting into the chat. It can have the whole project as context,plan features with me, etc. I use it a lot in a way that I cant see why I would go to programming without it right now. It takes time to adjust to it to learn how it works and all but it's worth it
The price of electricity where I am in California is pretty cheap for the energy itself -- I pay about $20/mo for generation -- but the cost for electricity delivery is absolutely fucking insane. It costs me $90 for "delivery" of that $20 worth of electricity.
Yes, I misremembered some things. Apparently Mono has more compatibility with .NET Framework (for instance 4.81) than dotnet (the current, modern recently released in version 10).
I mixed that up to mean that .NET Framework proper was released as open source, but that's unfortunately not the case.
