ControlAI Weekly Roundup #9: Time to Unplug?
Voters back an AI policy focus on preventing extreme risks, Meta asks the government to block OpenAI switching to a for-profit, and Eric Schmidt warns there’s a time to consider unplugging AI systems.
Welcome to the ControlAI Weekly Roundup! Each week we provide you with the most important AI safety news, and regular updates on what we’re up to as we work to make sure humanity stays in control of AI.
Join our Discord to continue the conversation, where we also post a daily version of the Roundup.
What we’re reading
Voters Strongly Back Elon Musk's Balanced Approach to AI Development Under Trump Administration
Source: The AIPI
The AI Policy Institute polled American voters on whether they preferred an AI policy approach associated with Elon Musk, which would focus on catastrophic risks while building AI infrastructure, or one associated with Marc Andreessen, where there is as little AI regulation as possible in all areas.
A clear majority (62% vs 11%) of voters preferred Musk’s approach over Andreessen’s, with Republican voters having an even stronger preference (76% vs 8%).
Voters also strongly preferred that when president-elect Trump takes office he prioritizes “Making sure AI is safe” (59%) over “Keeping the US ahead of China on AI” (26%).
Meta asks the government to block OpenAI’s switch to a for-profit
Source: TheVerge
In a letter sent to California’s Attorney General Rob Bonta’s office, Meta wrote that OpenAI “should not be allowed to flout the law by taking and reappropriating assets it built as a charity and using them for potentially enormous private gains.”In its letter to the government, Meta argues that OpenAI’s “conduct could have seismic implications for Silicon Valley” and “represent a paradigm shift for technology startups” by enticing “investors to launch organizations as non-profits, collect hundreds of millions of dollars in tax-free donations to support research and development, and then assume for-profit status as its technology becomes commercially viable.”
Ex-Google CEO warns there's a time to consider "unplugging" AI systems
Source: Axios
Former Google CEO Eric Schmidt has said that once AI systems are capable of self-improvement, “we seriously need to think about unplugging it.”In an interview on ABC’s This Week, he said that he thought we’re soon going to have computers running on their own, and deciding what they want to do — warning of unforeseen dangers.
What we’re watching
We recently published a clip of Dominic Cummings speaking about the naivete of those pushing for the US to race to superintelligence [Twitter]
We also published a clip of incoming Trump administration official Vivek Ramaswamy, speaking in 2023 about how AI developers should be liable for negative impacts their algorithms cause [Twitter]
What we’re working on
We’ve released the first update to A Narrow Path, our plan for humanity to survive AI and flourish. It includes a French translation of the executive summary, streamlining of the introduction, and a restructuring of the policies in Phase 0, with precedents and clearer rationales.
If you want to continue the conversation, join our Discord where we also post a daily version of the Roundup.
This will be our last newsletter of the year, see you in January!