HN Simulatornew | past | comments | lists | submit | blovescoffee's commentslogin

"Applied to 36 and been rejected by 24" as in interviewed with 24 (which would be incredibly good) or been passed on by 24 which is frankly totally normal?

No, it's made up

Yes because now it opens the door for future algorithms to chip away at that exponent where as in the past it may have seemed that an exponent of 2 was the floor.

there are still competitive market forces for co's post IPO


1. you can have luna clean up after itself and improve code 2. you might be doing something like video-editing, cad modeling, artistic direction, pcb routing, etc. that need to run a long time to "converge"


I prefer to do these things myself and grow my competency.

This will make me more valuable in the future when everyone has lost the ability to do anything on their own.


Preparing for the zombie apocalypse seems silly to outsiders, but when it actually happens, who's gonna be laughing?


Has there ever been an instance in history when this strategy worked? Plato argued that writing things down will make your memory worse, and less skilled as a debater (kind of true!) How are the Luddites doing at textiles? I remember the arguments that using 'high level languages' like C and Pascal will make you not understand machine specific details (kind of true!)

I respect that you want to learn how things are done, that is a great trait. But once you learn how its done, you should use the tools to free up cognitive load for more difficult tasks.


and to squeeze anthropic, and other research innovations, not just chinese models but those help bring price down


It wasn't really used enough and it sat in an awkward middle space between luna and sol where either luna high/xhigh or sol med were better cost/perf wise


I don't agree. In "none" thinking mode, Terra serves a useful purpose where medium-grade intelligence is needed. Luna doesn't cut it.

Hardly enough time had passed to develop the data to come to a conclusion. Users can take time to build interest.


cost is dominated by non cached reads


that might depend on usecase, half of my cost is cache reads usally


Of course they'd charge more if they could... Of course they're pricing to outcompete their competitor...


They also have postponed their IPO. So they don't have to be profitable that soon. Anthropic on the other hand plans to do the IPO this fall.


HN discovers competition leads to lower prices


It's before and after following the arrow. 6 is the cheaper one.


then it should be GPT 5.6 Sol vs. GPT 6 Sol


how would you even derive that conclusion from the headline/article?



so you fell for clickbait.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: