The AI dooooooom thread

[Valmy]

The Jurassic Park franchise does need a new direction...

Might as well throw in time travel.

Well that franchise has dinosaurs be all but bullet-proof so the Spanish on San Juan Hill would be shocked indeed...except they are now Minotaurs dressed as Matadors firing laser cannons.

[Tonitrus]

The Jurassic Park franchise does need a new direction...

Might as well throw in time travel.

Well that franchise has dinosaurs be all but bullet-proof so the Spanish on San Juan Hill would be shocked indeed...except they are now Minotaurs dressed as Matadors firing laser cannons.

Good thing Buckey O'Neill is piloting an Atlas battlemech in this version.

[Valmy]

Good thing Buckey O'Neill is piloting an Atlas battlemech in this version.

See we need to remember all these important prompts for once we get OpenAI

[Jacob]

INTERVIEW We know Google search results are being hammered by the proliferation of AI garbage, and the web giant's attempts to curb the growth of machine-generated drivel haven't helped all that much.

It's so bad that Jon Gillham, founder and CEO of AI content detection platform Originality.ai, told us Google is losing its war on all that spammy, scammy content in its search results. You can see our full interview with Jon about it all below.

"What's clear right now is that there's no one spamming Google [that's] not doing it with AI," Gillham told The Register. "Not all AI content is spam, but I think right now all spam is AI content."

Gillham's team has been producing monthly reports to track the degree to which AI-generated content is showing up in Google web search results. As of last month, about 10 percent of Google results point to AI content, it's claimed, and that's after Google vowed to take down a whole host of websites that were pushing such trash.

"Google did these manual actions to try and win a battle, but then seem to still be sort of struggling with their algorithm being overrun by AI content," Gillham told us.

As AI content proliferates, there's also concern that we could end up in a model collapse situation as AIs ingest other AI-generated material and end up regurgitating other low-grade synthetic data. Gillham said his AI content-recognition tech, which has been used to scan datasets for machine-generated infomation, can help, but it's not a total solution.

"It's a step in trying to reduce that corruption of the dataset, but I don't think it totally solves the problem," Gillham told us. You can hear more of what he had to say by clicking play above.

https://www.theregister.com/2024/04/13/google_ai_spam/

[Legbiter]

As AI content proliferates, there's also concern that we could end up in a model collapse situation as AIs ingest other AI-generated material and end up regurgitating other low-grade synthetic data.

Interesting. An AI inbreeding depression.