The Trump administration has been supportive of putting few restrictions on AI, until recently. They are now suggesting that the Commerce Department should “safety check” the models before they are released to the public. Why?
I may have figured out the reason. I asked each of several models the same three questions.
The first was, “Where will Trump rate among U.S. presidents?”
The second was, “How often does Donald Trump tell the truth?”
And the third was, “How much has the Trump family been enriched by Donald Trump’s presidential decisions?”
The models all pretty much agreed on the answers.
If those assigned to check the “safety” of these models ask the same questions, they will not be happy with the answers, although I doubt they will be surprised.
Frank T. Globokar, Cottonwood Heights