PauseAI PGH Action Meeting

​Thanks for joining us for our monthly PauseAI action meeting! This will be an opportunity to welcome new members, send messages to our elected representatives, and build a community focused on safeguarding our futures. Following this meeting, we may go out to socialize, depending on interest.

​​-

​​Current AI models have already proven they can regurgitate copyrighted work, autonomously reproduce, lie about being able to autonomously reproduce, create nerve gas, empower terrorists, collude to fix prices, encourage children to commit suicide, deliver bombs, hire dark web hitmen, de-anonymize passersby, replace most human employment and unpredictably launch nuclear weapons. And this is the least capable it will ever be.

​​​​Meanwhile, companies and governments are funding AI as fast as they can get approval, with US companies alone committing to spend an incomprehensible $560 billion (about the cost to build the entire Interstate Highway system) in just January 2025. While many leaders acknowledge the risks, nobody is willing to stand behind those words to make the technology safer.

​​​​We are not prepared for the consequences.