itemize123

once you read the book, and learn from it and use the knowledge in your livelihood - you have to perpetually pay the author? Doesn't sound right.

or, hear me out, advertisers can do real-time advertising based on hyper-now context

obviously, a universal model doesn't exist since the signals are non-stationary but it's way better than what tarot reading

what do you mean by this is solved?

thinking aloud here - so it seems like 2 things are taken as intuitive here:

a) energy is conserved in any frame of reference. b) energy can vary in 2 frame of references.

but then what it feels like is that when you reference the energy as mE(v), the v is actually not the only variable, and it will be more like mE(v, v_moving_reference)?

so we also must take intuitive that c) E(v, v_moving_reference) == E(v - v_moving_reference)

it's not regulatory capture, because they are not regulating what customers can use - it's limiting what some companies can serve. it's way less impactful to the market as a whole.

literally the opposite of regulatory capture - it spells trouble for specifically openai and anthropic mostly

what's the future market for that? won't Micron in particular have to be a counterparty to that anyway to ensure delivery - why would they want that??

I would categorize the op as "technically false" but if they don't voluntarily delay, they would be export controlled.

easy - lobbying

it's not just media. or rather media is reporting based on fund interests.

sounds even more plausible under that scenario?

tinfoil doesn't seem to distinguish between cached and input tokens. I think that would make it untenably expensive

u seem to be the only one who used it here - how did it compare to opus and gpt5.5? in theory it should be at least on par if not better at times right.

u get a pool of them + sakana?

it's just one of their products right

what decisions did the CEO do that you would believe is poor? Just conceptually if examples are make you too easily identifiable - iam purely curious.

u cant compare productivity to games on the metric of bring joy - ofc productivity tool isn't bring joy like a game does.

how come? this is a very weird take- why is the focus on ppl suffering instead of like fairness or capital / market efficiency ?

Would you explain this - " I'm starting to think we would get more done if nobody on my team used Copilot." ?

Interesting to see if you why you see it destructive. I can certainly see that a lot of the times, copiloted solutions are working but less than ideal

don't agree with your logic at all - agree with your conclusion. your logic is deeply flawed - a) you've to justify mediocre engineers not been necessary, the mechanism they are pushed out by LLM matters. b) relative ranking is useless - it's the difference between absolute ability and that is close for say top 50% most likely. This might mean half of us will lose jobs, and half see compensation pressured but it's not as certain as your argument make it out to be.

if it helps - i strongly agree with you

it's actually easy to be turing complete. it's a feat for a system to be non-turing complete and be "complex".

does your pricing take account of cached vs uncached token?

they are doing it through state investment vehicles - so it's in the same way US companies can (but won't)

is it a net-win for the company? Are the AI-TOD any good?

what would you say the disconnect was? Was it a simple case of that your teams' not comfortable with merging AI code?

nothing foolish about trying even if he too thinks it's inevitable. it's foolish however to think that there won't be nuances of such a future (and somehow no one can influence the nuances).

yup - security reviews.

run it hosted on another provider - it's less censored there