This is why a handful of corporations running essential government infrastructure and services (healthcare, welfare, comms, security) all over the world should make us all nervous..
SAFETY
-
Autonomous Vehicle Decision-Making in Complex Traffic Scenarios
By
–
this problem is incredibly complex.
— swyx 🇸🇬 (@swyx) 18 novembre 2022
here we are going in two lane traffic, but someone ahead of us has opened a parked car door on the side of traffic.
Our vehicle paused for them, waited for cars on our right to pass, and then *overtook* the human that was blocking us. pic.twitter.com/9J0nsnyfaBthis problem is incredibly complex. here we are going in two lane traffic, but someone ahead of us has opened a parked car door on the side of traffic. Our vehicle paused for them, waited for cars on our right to pass, and then *overtook* the human that was blocking us.
-
Galactica AI Tool: Promising Yet Risky Like AutoPilot
By
–
Great thread! I will try #Galactica with the same caution as Tesla's "AutoPilot." It's fun but dangerous. It should not be shut down. Great for writing onion articles!
-
ImageNetX: Human Annotations for AI Vision Model Robustness Analysis
By
–
We’ve released ImageNetX: a set of human annotations for the popular ImageNet benchmark to gauge model robustness strengths/weaknesses — one of the first large scale efforts to pinpoint mistake types in AI computer vision systems. Explore the dataset
-
EA Tweetstorm: Effective Altruism and AI Ethics Debate
By
–
i've been stopping myself from sending my EA tweetstorm for a week but idk how much more self-restraint i have
-
AI Safety and Capability Are Not Orthogonal Vectors
By
–
it's a dangerous myth that AI safety and AI capability are orthogonal vectors. also a myth that we can "avoid ruin" without careful iteration and contact with reality. things have not gone as our best experts have predicted, and that will continue to be the case.
-
Tabula-Rasa Model Oxymoron: NFL Theorem and Fundamental Assumptions
By
–
"Tabula-rasa model" is an oxymoron, by the NFL theorem. The question is what are the fundamental assumptions that need to be built in. Causality is a candidate. The laws of physics are another.
-
Testing Stock Prediction Models for Security Flaws
By
–
Testing stock prediction models for security flaws 📈 pic.twitter.com/XO9AQCglr1
— IBM Data, AI & Automation (@IBMData) 16 novembre 2022Testing stock prediction models for security flaws
-
Self-examination in AI science models for error detection
By
–
what would be cool with science models like Galactica would be to add self examination of the inputs like the coming Codex model, where the model can inspect it’s theorem proofs or essays and spot errors
-
Concerns about GPT alignment with human values and safety
By
–
"Finally, we are very concerned that this GPT could be unaligned with humans. This would be bad. We want this to be a nice GPT that deeply loves all humans and is always considerate and helpful. Thanks"