AN synthetic intelligence knowledgeable has warned that we’re about to witness the autumn of AI minions.
Professor Marek Kowalkiewicz revealed his issues at South by Southwest 2024 in Austin, Texas.
“We’re about to witness the bots*** avalanche,” he warned.
The knowledgeable used the time period bots*** as a play on phrases for a standard derogatory phrase.
He explains that this AI avalanche will come as a result of some bots creating content material on the web are slowly deteriorating.
AI algorithms like these utilized by ChatGPT fashions be taught from content material on the web.
As AI produces extra content material that is on the web, Kowalkiewicz warned that bots will begin studying from themselves.
He described this because the bots consuming themselves like a carnivorous and confused snake.
“There’s a lot of people that say it’s [generative AI] not that good anymore,” he told a small audience.
Even OpenAI has previously admitted that ChatGPT was acting lazier than usual.
“We’ve heard all your feedback about GPT4 getting lazier!” the company wrote on X last December.
“We haven’t updated the model since Nov 11th, and this certainly isn’t intentional. model behavior can be unpredictable, and we’re looking into fixing it.”
Kowalkiewicz added that generative AI is “eating its own tail” by “consuming its own content.”
The knowledgeable expects a downfall of AI bots will occur within the subsequent few years.
He suspects that generative AI fashions will run out of recent content material to be taught from on this time interval and begin consuming nonsense.
Kowalkiewicz gave the instance of bogus Google Search outcomes which can be already showing as a consequence of misinformation in AI content material.
An AI avalanche like this might have a adverse influence on the data we devour and the way properly we belief on-line content material.
The knowledgeable explores this concept additional in his new guide The Economy of Algorithms: AI and the Rise of the Digital Minions.