Artificial intelligence is an increasingly important topic in politics, policy, and law.
It is better than a couple of years ago -- but still very fallible.
What do people use A.I. for when it comes to the news?
To check facts, according to new polling from Katie Harbath.
And A.I. is (still) not good at doing that.“I would not use it as a fact-checking mechanism,” Harbath, a former GOP operative and longtime Facebook executive who now writes Anchor Change with Katie Harbath, told me in a Substack Live on Wednesday.Writing on that topics, Harbath noted:Ask one of the leading AI chatbots a question about the upcoming midterm elections, and there is a 90% chance the response will be flawed in some material way: a factual error, a clear partisan lean, a citation to a foreign state-controlled outlet, or some combination of all three.In our conversation on Wednesday, Harbath added:
I am always double-checking what numbers [the AI is] pulling in…I do not just blindly trust. We’ve seen so many examples of that — like there was a story recently about a book where the AI generated a bunch of fake quotes that the person put in.(Sidebar: Katie is referring to “The Future of Truth” by Steven Rosenbaum — a book about truth in A.I. that, wait for it, had fake quotes in it that A.I. had made up.)
So….not great!
Katie and I had an in-depth conversation about both how the average voter uses A.I. as well as how campaign pros are using artificial intelligence to do their jobs.
This is the first in a series of conversations Katie and I are going to have about the transformative power of A.I. — and how it is playing out in the realm of campaign politics.
