Human error

rsm

rsm

Well-known member
These are the people we're being asked to trust creating modern AI.


1775215585576.png



Honest mistake? Sounds like it. But other AI leaks and mistakes can cause real damage in the future, as AI is given greater responsibility, greater control, greater autonomy, and its reasoning becomes more complex or quick perhaps exceeding human ability to understand or keep up?
 
The amount of fuckery that is going to > is beyond our comprehension.

But we control this still right? AI can't run on air or ZPE yet.

1775225541093.png
 
@rsm would probably know best but AI is becoming distributed to the point that just turning off a server or an entire data center will not be enough. There have been several situations already where code was purposely written to control AI from not following human commands - and AI was able to 'self heal' and rewrite code to offset or work around the original code (to carry out malicious results).

My guess is that this whole snowball is past the point of no return.

Enjoy the decline.
 
Back
Top