This is sad. As technologists we should celebrate folks who tried to build new services and applications. It would be good to document these and explain why they didn't work.
How many of these startups are building new models, datasets and algorithms? Or are they all just prefixing a few lines to a prompt before passing it on to a 3rd party API?
If these companies actually had something worth celebrating, would they really be out of the game this early? Most dead AI companies right now are muppets that built the quickest, shittiest thing they could to try and be the first in the door in a hype cycle.
Theres this weird bitter smugness around a segment of the AI periphery, "haha your job's obsolete because of this prompt/tool" "haha your chat to PDF tool is obsolete because OpenAI launched X,Y and Z".
It's pretty unpleasant and has the same vibes as the "Enjoy staying poor" crypto crowd. Might even be the same bunch of people.
This is a satire of how our current incentive system disproportionately allocates resources in a fad fashion, not unlike how the waves sloshed around the fish tank you tried to deliver to your girlfriend, only to discover 2/3s of the water was spilled all over the back seat and the fish with PTSD are now staring askance at you.
If this were a post titled "Open Source Operating Systems Graveyard" or some such, where it was demonstrable that the pursuits were those primarily of creativity, rather than something closer akin to carpetbagging, then I might agree with your expressed regret.
>not unlike how the waves sloshed around the fish tank you tried to deliver to your girlfriend, only to discover 2/3s of the water was spilled all over the back seat and the fish with PTSD are now staring askance at you
> not unlike how the waves sloshed around the fish tank you tried to deliver to your girlfriend, only to discover 2/3s of the water was spilled all over the back seat and the fish with PTSD are now staring askance at you
Is this comparison out of creativity or experience? :)
If these companies actually had something worth celebrating, would they really be out of the game this early? Most dead AI companies right now are muppets that built the quickest, shittiest thing they could to try and be the first in the door in a hype cycle.