"On the creative product innovation task, humans supported by AI were much better than humans. But for the problem solving task, [the combination of] human and AI was not even equal to humans…meaning that AI was able to persuade humans…of something that was wrong."
SAFETY
-
AI Self-Verification and Reflection Capabilities Discussion
By
–
That’s incredible. Can you get it to check its work? I wonder how good self reflection will be.
-
Thermostats Control Without Mental States: AI Ethics
By
–
Thermostats implement control without mental states.
-
Controlling AI Systems Through Conversational Design and Balance
By
–
"If you really focus on creating an AI that's smooth, conversational, fluent, very even-handed and balanced, it turns out it's possible to do it. Pi very kind, balanced and friendly and that demonstrates that we can actually control these AI systems."
-
Denying AI flaws delays progress and innovation
By
–
this hectoring bullshit has got to stop. denying real flaws just delays progress.
-
Red-teaming AI Systems: Demonstrating Vulnerabilities Through Security Testing
By
–
fairish point, but if they ever let us red-team the system I am sure we can demonstrate something like this…
-
Sora’s Generalization Capabilities: Artifacts in Human Figure Generation
By
–
The completely amazing thing is how this baby generalizes!
— Gary Marcus (@GaryMarcus) 18 février 2024
I mean probably Sora has probably never been trained on people with three arms. (And what’s up with her neck?) https://t.co/xNBQmdqR3SThe completely amazing thing is how this baby generalizes! I mean probably Sora has probably never been trained on people with three arms. (And what’s up with her neck?)
-
Neural Networks Without Symbolic Representations Vulnerable to Hallucinations
By
–
Literally I have arguing for 23 years that neural nets without symbolic representations would be vulnerable to hallucinations, and they have. And literally every year (often daily) I have heard promissory notes like these, and never seen a detailed technical explanation as to https://
x.com/saehtweets/sta
/saehtweets/status/1759329436007112736
… -
OpenAI’s New Game: Spotting AI-Generated Video Violations
By
–
See how many physical and biological violations you can spot in this video. Fun new game, brought to you by OpenAI.
-
End-to-End Models and Neural Network Interpretability Explained
By
–
End-to-end just means that no part of the model is crafted by hand. The models don’t do anything incomprehensible. https://
distill.pub/2020/circuits/