I explain why trying to "ensure safety" actually hurts safety in this analysis:
SAFETY
-

Safety Challenges in General Purpose AI Models
By
–
The *big* problem here is point 4. It is literally impossible to *ensure* safety of a general purpose model, and attempts to do so are likely to *reduce* safety.
-
YouTube Requires AI Transparency Disclosure for Creator Content
By
–
YouTube is adding new AI transparency requirements for videos made w/ generative AI. https://
tinyurl.com/yc3k5f8u I've been curious if social platforms would update policies to require creators disclose GenAI content. (Could help w/ efforts to mitigate AI risks like deepfakes, etc.) -
LLM Safety Concerns: Why Major Risks Haven’t Materialized Yet
By
–
Because LLMs have been around for several years and no such thing has happened?
-
AI Team Shuts Down Project Over Ethical Concerns
By
–
Nope. It was taken down by the small team that built it because they couldn't sleep at night.
-
MBZUAI President Argues AI Existential Risk Concerns Unfounded
By
–
A piece by MBZUAI president @ericxing in WEF Agenda explaining why worries about AI existential risks are baseless.
-

AI Refuses to Acknowledge Common Knowledge About Mules
By
–
Then I also had to explain that sometimes the Knows Everything box is so unwilling to make anyone feel bad that it will pretend not to know the thing that all Americans know about mules.
-
Parent Ignores ChatGPT Safety Advice for Daughter Swimming
By
–
In this case I was trying to decide if it was safe for my daughter to swim. I decided to ignore ChatGPT and I told her no…
-
Google CEO Says AI Ethics Experts Essential for Development
By
–
Google's CEO now claims they need ethicists to work on AI
-

AGI Reduces Existential Doom Risk Compared to No AGI
By
–
P( Doom | AGI ) < P( Doom | no AGI ) #AGI #AGIFirst #ExistentialRisk