AJ Sadauskas on Nostr: And so the enshittification of ChatGPT begins. CMOs are now looking for ways to game ...

And so the enshittification of ChatGPT begins.

CMOs are now looking for ways to game those large language model algorithms.

After all, millions of people are now turning to ChatGPT for advice and uncritically accepting what it says as true.

So if ChatGPT says "product xyz is the best", they'll take that as gospel.

So a cottage industry of AI optimisation specialists is popping up on LinkedIn to take the CMOs money, and spam the sites the LLMs are trawling for training data.

That means spamming Reddit and Wikipedia, two human-edited sites that were among the best places to find information from an actual human on the web.

Which is why they're weighted so heavily in the ChatGPT and Google Gemini training data.

And now with a cottage industry of AI optimisation experts trying to game AI responses, they're sbout to be spammed to kingdom come.

See the LinkedIn post below for an example.

So now not only are you going to be increasingly likely to see AI queries return responses about how you need a Coke.

But legit information sources will be spammed to make it happen.

#LLM #AI #artificialintelligience #ChatGPT