Anthropic pitches AI safety bill that may block a model if 50+ people could be killed
Published
What's this about?
The proposed policy seems to offer protections against AI run amok but actually forbids federal agencies from blocking a model's release without the permission of a judge.
The one-panel cartoon
Read the cartoon's script
A high-tech meeting room where AI safety experts and lawmakers are discussing the bill, with screens displaying flowcharts and warning signs.
Caption: When AI safety means getting a judge's permission to halt an apocalypse
Read the full 3-panel strip
Based on the original story on techcrunch.com.
Editor's note
As we navigate the uncharted waters of artificial intelligence, it's crucial to balance innovation with safety. This proposed bill, while well-intentioned, highlights the complex legal challenges in ensuring AI's responsible development without stifling technological progress.
Rate this comic
Read the script
Panel 1
A small meeting room at Anthropic, with a whiteboard and several chairs. The room is modest, with posters of AI ethics principles and a few plants. Characters (Ella) *Ella:* *Max:* *Ella:*
Characters: (the AI safety officer) and *Max* (a junior developer).
Ella: So, our bill would require a judge to sign off if we think a model might cause harm.
Max: Wait, does that mean we need a judge’s approval just to issue an update?
Ella: Yeah, but it’s only if there’s a serious risk...
Panel 2
The same meeting room, with the team looking at a mock-up of the proposed bill, faces slightly exasperated. *Ella:* *Max:* *Ella:*
Ella: It’s supposed to protect us from overreach, but it’s turning into a red tape nightmare.
Max: Like, what if a model malfunctions and could kill 50 people?
Ella: Even then, we’d need to prove it to a judge first...
Panel 3
The same room, but now Ella and Max are looking at a screen showing a mock trial with a judge and lawyers. *Ella:* *Max:* *Ella:* *sighs*
Ella: And the judge would have to decide if we should release the model or not.
Max: So, if our model might cause harm, we need a judge to say it’s okay to hold it back?
Ella: Yep, and it’s supposed to protect us from government overreach...
Highlight: The absurdity of needing a judicial review to prevent a potentially dangerous AI release, when the goal was to prevent unnecessary government overreach.
Comments