repparw
repparw on Hacker News
a unicorn like you're mentioning isn't linear. a 1% 1000x business probably can stop along the way when you reach a TC you're comfortable with, that you have a 1% of hitting that 1000x doesn't mean you don't have a 25% of hitting 2-4x with the same idea before.
There's a lot of people who made their money by bootstraping business 0 to 100k, etc, and selling. (There's also people who center on buying 100k business and growing it to the million stage... and so on)
I would test this, might be cheaper per task even costing more per token, probably faster too
> once I get the business going I'd likely begin to buy my own machines
only when it's cheaper than to continue renting it in your calculations
same goes for renting vs buying a house, or anything.
the breakeven point here would come when they stop subsidizing the subscriptions, or when local ai gets to run on basically everything, making it ~free (sans electricity)
some people need "more calories" than others
fair != equal
if you care for some feedback about the writing, dropping the link and saying "PR's are open!" would land probably equal or better, and would reduce noise on the message. as sibling said, substance and noise
we're reading comments on a post about math proofs
and if it goes fast enough, it can fail and you can prompt again before you need to worry about if it's going in the wrong direction. or, it will try multiple directions!
anecdotally, a lot of argentina's apps that have a similar system "had" to develop an interop with brazil's PIX as a feature, ideally an endgame would be for these systems (wero, pix, other countries') to all standardize and integrate each other. would love something like that.
Didn't we get this with Harry Potter back in like gpt3.5? I'm sure I saw some news about it, someone getting it to output a book's intro word by word, couple pages?
the engineering workarounds to give the Neo a second port probably says that their internal numbers differ, regarding the number of ports importance
- Not enough compute for the requests they have
- Selling those requests at less money than it cost to run the compute for those requests (because if you raise price clients go to openai)
The statements are not contradicting each other?
They keep subsidizing to try to grow customer base, but they can't serve the customer base they have, they're expecting customer base grows faster than it drops from people bothered with rate limits (it probably will, average user won't hit rate limits enough to change)
Probably expecting a breakthrough in efficiency for compute, or getting enough cash flow (IPO?) to get more compute before it all comes crashing down
Could the inverse process work? Design, then (crowdsource?) pay for UL certs?
Don't know the process for certification, but in the software side, I've heard from foss projects asking for money for, e.g, security audits
Motorola was the only one that had something similar AFAIK (Moto's Ready For)
Though I'd expect that all efforts focus on the new Android Desktop Mode now, and then Samsung Dex turns into something akin to what OneUI does with Android, instead of being its own thing
Did you get one? I'm looking for one too, search got me here
judging by the .ar ccTLD he's from Argentina, same as me.
First hand experience tells me local ISP's don't care, and/or don't know to care. they don't even serve piracy notices here (I believe most of latin america is like this) so they definitely won't be bothering with something like this