I appreciate Claude 3's ability to roll with whatever weird AI generated images I give it …until you hit its odd guardrails around human-wildlife interaction: "the idea of casting dinosaurs, especially predatory ones, in an action film alongside humans raises ethical concerns"
SAFETY
-
AI takeover of human consciousness at early age
By
–
I think you took over your host's brain when the host was 4½
-
Neuralink Security Breach Risk Raises Critical Safety Concerns
By
–
Why did nobody bother to check if the Neuralink device was compromised? I just hope Noland is ok and the bot did not get into his brain
-
ChatGPT Android App Adds User Control for Voice Data Training
By
–
ChatGPT Android app now has an option to allow Open AI to use your voice data for training purposes. The option is disabled by default
-
AI-Generated Images Used in Facebook Spam and Scams
By
–
Are you noticing more viral, bizarre photos on Facebook? AI-generated images are now being incorporated into spam and scams since the technology became available. A new preprint paper from HAI affiliate Renee DiResta (
@noUpside
) examines the phenomenon: -
GPT-4 Demonstrates Superhuman Persuasion Abilities in Controlled Study
By
–
AI is already capable of superhuman persuasion In this randomized, controlled, pre-registered study GPT-4 is better able to change people’s minds during a debate than other humans, when it is given access to personal information about the person it is debating.
-
James Zou on Interpretable and Trustworthy AI
By
–
James Zou, Assistant Professor of Biomedical Data Science at Stanford University, provides a deep dive into the realm of interpretable and trustworthy AI.
-
AI Existential Risks: Psychological Denial and Family Protection Concerns
By
–
I think it's a nightmare, with oscillations into denial, and the coverup is mostly managed to protect the lives of one's family members
-
Top AI Leaders Share Insights on Future Possibilities and Dangers
By
–
Curious about the future of AI? @BusinessInsider rounds up insights from top AI leaders like @StanfordHAI co-director @drfeifei
, distinguished education fellow Peter Norvig, and faculty affiliate @AndrewYNg
. https://
businessinsider.com/ai-godfather-t
op-names-possibilities-dangers-openai-chatgpt-list-2023-8
… -
Testing Shows AI Product Fails to Meet Expectations
By
–
So many people are testing it and reporting that it doesn’t quite work. Which is exactly the point I was making!