Comfortably Numb: Yo dawg, we heard you liked ChatGPT, so we put ChatGPT into your ChatGPT.

Yo dawg, we heard you liked ChatGPT, so we put ChatGPT into your ChatGPT.

This will only work if CriticGPT is AI in the same sense Amazon cashless stores are AI www.theguardian.com/commentisfre...

So, Amazon’s ‘AI-powered’ cashier-free shops use a lot of … humans. Here’s why that shouldn’t surprise you | James Bridlewww.theguardian.com This is how these bosses get rich: by hiding underpaid, unrecognised human work behind the trappings of technology, says the writer and artist James Bridle

You know what? If they'd pay me money to rip apart garbage AI output, i'd take it.

Oh boy do I have a few choice words about the code I have seen it generate.

I bet that the corporate plan is to sell products with "No AI was used" sticker somewhere down the road

I can put you in touch with someone who’s hiring for that. It’s not as fun as it sounds and pays shit.

ChatGPT has decided that ChatGPT will use ChatGPT to check ChatGPT for errors, says ChatGPT

Sounds like ChatGPT has been learning from law enforcement.

Not enough passive voice

FibonacciGPT

It's working great tho

What could possibly go wrong?

A cascade of errors

Who did quality assurance?

MARC JACOBS

But who actually makes it?

Now with double the energy usage!

Giving Spam, Spam, Spam, egg and Spam

Without the egg...

The circular irony of chat gpt fixing chat gpt

We just need CriticCriticGPT to fix CriticGPT

HecklerGPT can just make jokes about. InsufferableAssGPT can deliver lectures to us for every error the Heckler heckles.

All I can think about is how much more the planet will have to burn to correct the original planet burning program's mistakes O_O

"ChatGPT has investigated itself and found no misinformation"

how might one debug this infinite hallucination loop nightmare?

Can we just replace CEOs with LLMs, which might actually save money and limit downsides

Easy. The only job LLM can really replace bsky.app/profile/numb...

We can replace him. We have the technology.

ChatGPT all the way down

Which one is the slave wage worker from an impoverished region?

OpenAI announces ChronosGPT: Intelligent AI Model sent back in time to stop errant sentient fork in CriticGPT no worries things should be back normal by two weeks ago

Put AI on your AI with a big red AI on top

Sort of like when they start opening Starbucks inside other Starbucks

The critic it stinks dot gif

So, here we are in the era of Beyond Silly. I don’t like it so far.

It’s like when everything suddenly had Bluetooth but with the bonus of destroying the planet

Qui code-stodat code-stodes..

(Or recursive GIGO)

Douglas Hofstadter not available for comment

We're all gonna keep laughing until this creates the Geth or some shit. (Google it)

This sounds a lot like grading your own homework.

Omg lol 😆

It's as if Tesla tried to bolt on a NN to control the car to improve its decision making on top of all the other NNs doing recognition and prediction. Oh wait... Now not only does it do stupid shit, you now also don't know what logic caused it!

In related news, ChatGPT proves P=NP Complete 😉

The critic catching ChatGPT's mistakes:

ah, so they’ll be employing TWICE as many input graders in third world countries

Now we have AI correcting AI? So, nothing bad will come of this, right? RIGHT?? 😱

I had to read that headline three times

Post