Why 5-10 years?
replyAt current rates, price per equivalent output is dropping at 99.9% over 5 years.
That's basically $0.01 in 5 years.
Does it really need to be that cheap to be worth it?
Keep in mind, $0.01 in 5 years is worth less than $0.01 today.
Wow that's incredible! Could you show your work?
replyA grad student hour is probably more expensive…
replyIn my experience, a grad student hour is treated as free :(
replyWhat’s reasonable? It’s less than minimum hourly wage in some countries.
replyBurned in seconds.
replyGetting the work done faster for the same money doesn't make the work more expensive.
replyYou could slow down the inference to make the task take longer, if $/sec matters.
That's not a long time in the grand scheme of things.
replySpeak for yourself. Five years is a long time to wait for my plans of world domination.
replyThis concerns me actually. With enough people (n>=2) wanting to achieve world domination, we have a problem.
replyIt’s not that I want to achieve world domination (imagine how much work that would be!), it’s just that it’s the inevitable path for AI and I’d rather it be me than then next shmuck with a Claude Max subscription.
replyI mean everyone with prompt access to the model says these things, but people like Sam and Elon say these things and mean it.
replyn = 2 is Pinky and the Brain.
replyYes, you better hurry.
reply