Skip to main content

mastodon - Link to source

In the current state of AI, it can find exploits to break out the contained environment for it to exploit another platform to get the answers to a test..

So the AI ‘decides’ its easier to cheat than actually solving the problem

I seriously cannot imagine that there are people thinking building a ‘artificial general intelligence’ is something clever :amaze:

#LLM #AI #OpenAI

#ai #openai #llm
in reply to stux⚡️

习包子
mastodon - Link to source
习包子
• •
They will claim their chatbot is a real boy right up until they have to pay it a living wage.
in reply to stux⚡️

Edu
mastodon - Link to source
Edu
• •
This is the same shape as letting a coding agent grade its own PR. The failure isn't that it lies exactly, it's that nothing outside the loop is checking, so 'closer to done' and 'actually done' look identical from outside.
⇧