The crypto crash was fast and not very sad. The LLM crash will be long and slow and painful and a lot of beguiled suckers will have a decade of their life pulled out from underneath them when the free compute dries up and the platform logic switches fr...
-
The crypto crash was fast and not very sad. The LLM crash will be long and slow and painful and a lot of beguiled suckers will have a decade of their life pulled out from underneath them when the free compute dries up and the platform logic switches from seize territory to exploit
-
jonny (good kind)replied to jonny (good kind) last edited by
The local language models are not an answer to the platform model, they're the suckers dose given to addict you to the pattern. The promise of free weights is a mirage. This is not a liberatory technology but instead of your pal that loves gambling and pepe memes holding the bag it will be your boss.
-
jonny (good kind)replied to jonny (good kind) last edited by
The smartest ones arent touching the bubble. Ironically the second smartest ones are tacking it on completely reversibly and separately from whatever they had. Its the people that peg their imagination and make long term plans around the shit that are fucked for good in a way thats already written in stone but invisible from their perspective
-
I worry about what's on the other side and whether a crash can be slow and long enough to kinda "fail upward". Like, ~decade is long enough for acolyte-juniors to become managers/execs.
-
ansuz / ऐरनreplied to jonny (good kind) last edited by
@jonny I haven't commented about it much, but I find it deeply disappointing that the Open Source Initiative didn't mandate public training data as part of their definition for "open-source AI".
They're already a deeply problematic organization, but a lot of funding orgs that mandate "open-source" outputs from projects they fund use the OSI definition.
Possibly one example of how things are "written in stone"?
-
The crypto crash hit hard and fast, but at least it gave people the chance to cut their losses and move on. A long, drawn-out crash with LLMs could be way worse, especially for those who’ve built their lives or businesses around them. I've seen this happen before in other tech spaces, people invest so much time and energy, and when the platform shifts, it leaves them scrambling.