Did it the day before yesterday with a fresh win11 image from Microsoft (no rufus magic or anything). You can press shift + f10 and run a command I don't remember. It restarts the install process and lets you create a local account while it still patches the system before the first start (no fiddling with internet connections required). At least that still works.
I see many people saying deepseek and other chinese providers have always been profitable. Also they show their training costs publicly. Can't say for sure since I have not used it personally, but I think they'll for sure outlive the western SOTAs.
OpenAI apparently runs a profitable inference business with 40% gross margin, but their advertising budget is nutso and their real costs are pretraining and research. I suspect Deepseek's comp is not predicated on capturing the lightcone of all future value, some googling insinuates their top pay is $212K US which would support that suspicion. Compare and contrast with the $1.35M and up at OpenAI.
> These are usually from a missing context (i.e. specifying where the car is right now).
That shouldn't be an issue. The very act of taking the car to car wash contains the information that it is with you and you are at your home. Doesn't this prove that LLMs don't actually think?
What if your car is not brought to the wash by you (common for companies or rich people). If someone asks you how to get it and the car wash is close by, of course you will tell them to go on foot. Human life experience is not universal and thus context can never be either. This is also a huge problem in inter-human communication, since most people suck at putting themselves in other people's shoes. But for LLMs everyone somehow assumes the humans must be right by default, when in fact you can tell even less about a model's training and world experience than for an average human.
