Samuel Hammond and Kelsey Piper debate AI timelines and lab lobbying
Samuel Hammond is defending a "realistic timeline" view of AI risks, citing leaked agentic AIs, a vulnerability explosion, and Anthropic's agentic malware report as already-realized harms. Kelsey Piper separately predicts Anthropic's DC lobbying will prioritize protecting itself over RSI warnings.
All posts · 5, most important first
- "Anthropic's lobbying in DC will mostly be focused on keeping Trump from messing with Anthropic and will not be in line with what the technical researchers are saying about urgent action on RSI being needed."1 likes · 0 reposts
- We've had several weakly agentic AIs leak onto the internet, are currently dealing with an explosion in vulns require partioning large swaths of the internet for security patches, and Anthropic just released a report on agentic malware. Another big theme of Leviathan is agentic flooding, which is also now happening. What were your 2023 projections?0 likes · 0 reposts
- This is not a doom post. This is my attempt to describe a realistic timeline which naturally includes both good and bad things. Technologies are packaged deals0 likes · 0 reposts
- Also note I didn't make any claim about the scale of damages. You are substituting in your own straw claim due to mood affiliation1 likes · 0 reposts
- Some obvious examples. Starts as slop content and expands across domains with capability and autonomy.
0 likes · 0 reposts
@kelseytuoc press · 
@hamandcheese policy ·