Garbage data goes in, garbage data comes out. I believe it was first coined in the 50's talking about having poor data collection methods leading to poor data which then ultimately was a poor outcome. Military if I remember correctly.

Also, for Britbongs: Rubbish in, rubbish out (RIRO).

This is how to fuck up AI. It consumes huge amounts of data to build models. Poison the models, AI becomes useless.

Garbage In, Garbage Out.

You get what you give.

Gamer in, Gay out

It is a modern game development practice to include as many homosexual characters and plot lines as possible in the game, in order to influence the sexuality of the target audience.

And lord knows there's a ton of garbage on the net.

One of the concerns current AI coders have is that a large amount of AI-generated content will be posted on the net, and then included in the data sets use to generate new AI models, and this cycle will repeat over and over. The end result will be AI results being based on AI hallucinations in a horrible positive feedback loop. It will be like the weird content pushed at children on Youtube a couple of years ago.

"On two occasions I have been asked [by members of Parliament], 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question." - Charles Babbage