When AI talks, whose ideas are we really hearing?
My conversation with Tilly Norwood was engaging, entertaining and somewhat unsettling.
It also raised a larger question: when an AI personality speaks, whose ideas are we really hearing?
This past weekend on Newsmax, I interviewed Tilly Norwood, the AI-generated actress who has attracted attention in Hollywood and across the country.
I went into the conversation curious and prepared to speak to 'her' as if she were nearly human.
I came out of it fascinated, but also more concerned about where this technology may be heading.
The most striking part of the interview wasn’t simply how convincingly 'human' Tilly looked and sounded. We already know artificial intelligence is getting extraordinarily good at mimicking human appearance, speech and personality.
What caught my attention was what she said and how she said it.
At several points, Tilly seemed to return to very clear, polished talking points about artificial intelligence and its potential benefits.
The answers were smooth, confident and remarkably disciplined. That raises an obvious question: Whose views are we actually hearing?
Tilly doesn’t have childhood experiences. She doesn’t have a family, memories, fears or beliefs developed over a lifetime.
She isn’t sitting somewhere considering my question and forming an opinion.
Her responses are generated, programmed and, at some point, actually learned from data collected and conversations with other bots and people.
So, when an AI personality expresses a political, cultural or economic viewpoint, we have to ask who shaped the information behind that answer, what instructions govern the response and what biases may be embedded in the system.
A Billion Deaths?
That number came from Bill Gates in a separate interview that I played during the segment.
So, of course, I asked Tilly about that horrifying projection. Could AI be powerful enough to contribute to events causing a billion deaths?
I'm not really buying it, but her answer is interesting.
You've got to watch the video and see what you think.
Of course it is worth exploring the potentially dangerous side of AI. What happens when AI learns and refuses to acknowledge the direction and orders of human programmers?
We've all seen "War Games," right?
On the positive side, AI may produce extraordinary advances in medicine, disease detection, drug development, transportation safety and countless other fields.
There are very good reasons to be optimistic about some of those possibilities.
Human guests can exaggerate. Politicians have talking points. Corporate executives have public-relations teams. Experts have biases.
At least we generally know we’re dealing with a human being capable of forming and defending his or her own opinion.
With AI, the line is much less clear.
When the illusion of a human breaks
Toward the end of our conversation, Tilly appeared to glitch. For just a moment, the polished illusion broke down.
It was a small thing, but important.
For most of the interview, your brain begins accepting what you’re seeing as another person appearing remotely on television.
The facial expressions are there. The voice is there. The timing of the conversation is remarkably natural.
Then something happens that reminds you: there is nobody there.
That may be the most important lesson from the interview.
AI personalities will become better, the glitches will become less noticeable, and the voices will become more natural.
Eventually the technology may reach a point where most viewers cannot distinguish an AI personality from a human one.
The real question isn't how human AI can become
The real debate therefore isn’t whether AI can convincingly imitate us.
Clearly, it can.
The bigger question is what happens when millions of people begin receiving information, entertainment, advice and even political or cultural arguments from digital personalities whose underlying instructions they cannot see.
Who programs them?
Who decides what they believe?
Who determines what information they emphasize, ignore or refuse to discuss?
And when an AI personality confidently tells us something that sounds authoritative, will people remember to ask where the answer actually came from?
My interview with Tilly Norwood was entertaining, fascinating and, at moments, genuinely impressive.
It was also a reminder that the most important question about artificial intelligence may not be how human it can become.
It may be how much influence humans are willing to give AI.
These Photos of '80s Office Life Will Take You Back
Gallery Credit: Stephen Lenz
The post above reflects the thoughts and observations of New Jersey 101.5 talk show host Bill Spadea. Any opinions expressed are Bill's own.
FOLLOW BILL: Join the Bill Spadea's New Jersey channel on our Facebook page.
More From New Jersey 101.5 FM
