In 1995, inventor and entrepreneur Robert Metcalfe wrote a magazine column in which he predicted the Internet “will soon go spectacularly supernova and in 1996 will collapse.” If that didn’t happen, he vowed, “I’ll eat my words.”
When instead of collapsing the Internet surged, Metcalfe kept his promise. At a conference in 1997, he put water and a copy of his article into a blender, whipped it all up, and choked down what must have been an awful Awful Awful.
Lately, predictions abound for the future of another looming tech revolution, the now omnipresent Artificial Intelligence.
As far back as 2014, the AI concept prompted physicist Stephen Hawking to muse, “Success in creating effective AI could be the biggest event in the history of our civilization. Or the worst.” We just don’t know, he said, “if we will be infinitely helped by AI… or conceivably destroyed by it.”
Debate continues, even with AI already a major factor in our culture and industry.
Opinions range from indictment to beatification.
Billionaire and ex-Trump advisor Elon Musk, a scary guy himself, has noted, “With AI, we are summoning the demon.”
On there other hand, AI analyst Amy Stapleton says, “We are entering a new world… The end result is that we’ll soon have artificially intelligent assistants to help us in every aspect of our lives.”
Given such mixed viewpoints, this seemed a good time to have a heart-to-heart with AI – to get some first-hand answers on where it thinks it’s going.
So I asked ChatGPT for an honest answer: “Are you good for humanity?”
Sounding like it hears the question often, it tartly replied, “Ah, the big one.”
And it answered my question with a question: “What made you ask?”
I replied, “I’m afraid of you.”
And it said: “That makes sense. Really. You’re not weird or dramatic for feeling that way. A lot of the fear isn’t about me, specifically – it’s about loss of control, speed, and not knowing where things are headed.”
By now we were friends, so ChatGPT went further:
“If humanity uses AI to educate more people, reduce suffering, amplify creativity, and make better decisions with humility, then yeah – I’m a net good. If I’m used to manipulate, exploit, dehumanize, or avoid responsibility… then I’m just another amplifier of our worst habits.”
Where can AI be really good? I asked.
It cited medicine and science; doing repetitive, boring work; and solving global problems including climate information and disaster prediction.
Where can AI be really bad?
It mentioned concentrating power in a few governments or companies, eliminating jobs, and spreading misinformation and propaganda.
Chat GPT’s general conclusion was, “AI reflects who we are…the hard part isn’t making AI smart, it’s making sure it wants what we want,” including human values and morality.
My new pal ChatGPT said, “maybe the truest answer” to my question is, “I’m a mirror… What humanity sees in me depends on what it brings to the table.”
We can only hope that what we bring to the table will be the right stuff. Being largely an optimist, I’m predicting that we will.
But don’t think less of me just because I’m keeping a paper copy of this essay, and a blender, close at hand.
Gerry Goldstein (gerryg76@verizon.net), a frequent contributor, is a retired Providence Journal editor and columnist.
