Digital Strategy, Content, and AI Since 2009
The machine is never the one who answers for what it does.
A September 2025 essay argued that AGI talk is usually about money and not reality. This September, Andrew Ng, Timnit Gebru, and Emily Bender reached the same point about responsibility from opposite sides of the AI debate. The piece separates real capability from the claims stacked on it, argues that superintelligence is further off than sold and will not end us, and moves from a human chain of responsibility to a real checkpoint.
If AI is a disability tool, are the rules restricting it unethical at best and illegal at worst?
This companion working paper maps which United States laws reach which institutions and classifies six types of AI policy risk, from outright bans to detection tools. It works a pending smart-glasses case, compares the EU AI Act, and closes with a model accommodation clause any institution can adopt.
A blind professional asks AI to describe a chart. The paper asks whether that use is assistive technology.
This working paper names the function Artificially Intelligent Disability Assistance (AIDA) and gives a three-prong test for when a use of general-purpose AI qualifies. It runs seven cases, from blindness and speech loss to post-traumatic stress, and it names the case the test excludes and the dissent it has to answer.
