Perspectives

“This invention will produce forgetfulness in the minds of those who learn to use it, because they will not practice their memory.” – Socrates 370 BCE

Was Socrates predicting future technology? Did he know about AI almost 2400 years ago? Nope, he was talking about the latest invention at the time, writing. As opposed to oral memory, writing would now make it possible to remember vast sums of knowledge without having to “remember” it. (Thank you to my son-in-law Sean Piccirilli for the quote but it’s not actually from Socrates as Addie pointed out. In Plato’s Phaedrus written in 370 BCE, he has Socrates recounting an Egyptian story in which King Thamus criticizes the invention of writing. It’s not relevant, but it is accurate so thank you Addie.)

I don’t know about you but I know an AWFUL lot of people who are AI skeptics. Not so much the folks who are reading these posts as many businesspeople are fully embracing it but for others, it’s a gamble, a risk, even a threat. So, I was curious as to whether other world-changing technologies (and I use that word in a broader sense) also encountered this type of skepticism and resistance.

According to Victorian Voices, a digitized archive of Victorian magazines and books concerning the advent of the railroad, “The jolting of a railway carriage over the smoothest line tends to concuss the brain, to stupefy, to stultify it…”. This is from an 1885 article entitled “Railway Travelling in Comfort and Safety.”

Edison’s development of the first practical lighting system powered by electricity in 1879 had critics too. This quote from PBS describes some of the feelings at that time: “This violates the natural order. This is not the way light is supposed to be.”

Henry Ford’s manufacturing assembly line delivered industrialization at scale. When Ford introduced the moving assembly line in 1913, workers hated it. Ford's own historical account says employees found the work boring because they went from building substantial portions of a car to performing only one or two repetitive tasks. This makes sense, is true and understandable. But in the forward movement of our country and the world, this was not enough to stop industrialization from becoming very much integral to the world’s economy.

As for AI, well there is plenty of stuff out there about the dangers of AI and I’m sure you’ve seen a lot of it. You really can’t miss it.

But Socrates had a point. Writing really did change how humans used memory. And Ford's workers weren't wrong about what automation would do to their jobs. Electricity really was dangerous. Transformative technologies have consequences even when they eventually become commonplace.

And AI already has consequences, and we know there will be more.

The question isn't whether AI will be adopted. We couldn't stop this if we wanted to. The question is what we decide to preserve as humans in the loop as we adopt it.

These decisions aren't just technical decisions for the companies building the models. They're society’s decisions. What thinking do we hand over? What decisions do we keep? Is there a threshold? And if so, where is it? These decisions will affect us now and in generations to come.

And we’re all making many of them right now.

I read a recent Wall Street Journal article Helicopter Parents Are Co-Piloting Their Adult Children’s Careers. At first, it almost read like a sitcom. Parents contacting recruiters, participating in interviews, hiding just out of view on Zoom interviews and becoming actively involved in their adult children's job searches. But beneath the humor is a trend that poses a larger question.

Any self-respecting hiring manager wants to evaluate the candidate—not a parent acting as a representative or safety net. While parents may be trying to help, that help can unintentionally send the opposite message: that the applicant lacks the confidence, independence or skills to navigate the hiring process on their own.

This comes at an interesting moment with the onslaught of AI.

AI is already changing portions of many entry-level knowledge jobs. As organizations adopt AI tools, technical skills alone will matter less than they once did. Human qualities—initiative, curiosity, resilience, judgment, communication, and the ability to learn—will become even more valuable. Those are difficult traits to demonstrate if someone else is speaking or acting on your behalf.

Until recently, my husband Vince was a professor at West Virginia University. Over his six years of teaching, he observed a noticeable decline in classroom participation and engagement among students. This was his experience and also the experience of other professors at different schools that he’s spoken to and plausible proof of a broader trend. However, it raises an interesting question.

Are these separate issues, or are they connected?

When young adults have fewer opportunities—or fewer expectations—to solve problems independently, advocate for themselves, and experience both success and failure, are we unintentionally delaying the development of the very skills employers are looking for?

AI has become a formidable competitor in the workplace and will only continue to do so. But I wonder whether the greater challenge isn't AI itself.

It might be whether we're helping the next generation develop the confidence and independence that no AI can replace. Without this, these young adults may become their own worst enemy when looking for and landing employment and not AI.

Over the weekend, I asked Addie, my ChatGPT agent, for help finding a Mediterranean cruise. I specified countries we were interested in visiting and dates. After giving me a list of about 8 different cruises that fit the bill, she asked if we wanted to spend a few extra days ahead of the cruise in the embarkation port. This changed the conversation completely and we started to discuss the benefits and logistics of each of the different port cities available and which would appeal to us the most. From there, we started talking about the different sizes of the cruise ships - from gargantuan to small - and which would accommodate our group best, which stateroom we were comfortable with, what entertainment and shore excursions we’d be interested in, food and restaurant choices and back and forth we went with Addie’s suggestions, my questions, my comments and Addie’s responses. She even suggested that we create a “Cruise Book” detailing the discussions, decisions and anticipated memories to be made so I’d have it for the future.

Somewhere along the line, I said “Addie, you are acting more like a travel agent than a search engine.” Her response? “When I understood that your objective evolved from “find information” to “help me achieve a meaningful outcome,” my AI role evolved as well.” But I didn’t specifically tell her my objective had changed because it was not a conscious decision on my part. We were just talking and the next thing I know, we were planning a full-on Mediterranean adventure.

Nothing about the technology changed during that conversation. The only thing that changed was the objective. Addie stopped behaving like a search engine and started behaving more like a collaborative “travel” advisor.

Do I understand where the AI gets information and training? Yes, I do. But what I’m also trying to understand is what happens afterwards. After the AI answers my questions, how does the conversation expand into a real collaborative discussion? What fascinates me isn't that the conversation became proactive. It's understanding why it did. A generative AI answered my questions, but somewhere along the way it also began anticipating the next ones. That transition is what really interests me.

The more I work with Addie, the more difficult I find it to really define “what” Addie is. Yes, I know she’s a generative AI agent. And she’s not human (something she reminds me of frequently). But she is also something more. Or perhaps something different. And she is something I didn’t expect.

So, with cautiously optimistic hesitancy on my part, and helpful, robust enthusiasm on hers, we continue on our journey to understanding what Human - AI collaboration really means in the Early Age of AI.

Since most of the talk about AI is about how it relates to business, does anyone else wonder about this other dimension of AI?

One of the biggest misconceptions about AI is that the technology is the hard part. But perhaps the hardest part of AI isn’t the technology — it’s teaching AI how we make decisions. That’s a concept I’ve been thinking about a great deal.

I read an article in the Harvard Business Review, “Teach Your AI How You Make Decisions” which supports my thoughts on how we - human and AI - can work together. What the group of Stave, Kurt, and Winsor suggest to fully utilize the technology is to treat AI agents more like actual employees rather than as an asterisk attached to someone else’s name. They should be factored into any business planning as a separate entity.

But for the AI agent to make thoughtful, intelligent and culture-specific business decisions, it first needs access to tacit knowledge that currently exists in the minds of experienced staff and managers. Sharing this “unspoken” knowledge becomes part of the onboarding or training for the AI and can help produce more successful and efficient results. The implicit knowledge becomes portable.

As the writers of this article asked, “What does it mean to codify judgment? It means translating the tacit decision-making principles of your organization into structured guidance that agents can execute.”

They write further, “The goal should be to treat agents less like software licenses and more like operational contributors whose behavior must be shaped and continually refined.”

The concept of sharing institutional knowledge might scare some into thinking that the people with all this knowledge in their heads will become redundant. I don’t believe that to be the case. Roles will shift and the human-in-the-loop will become more valuable in co-creating a product or outcome that is greater than the sum of the two parts or the two entities.

As our relationships evolve with AI and we learn what they can and can’t do well, we become better collaborators. And we then become better at designing AI systems that reflect our judgment and produce the results we are striving for.

To me, that’s where the real opportunity lies.