Reader,

I had dinner with my friend Mark last week and he asked me a question that seems to be on a lot of minds lately: “What do you think of AI?” And my answer also seems to be a prevailing answer: “I don’t know.”

We talked about the potential good and (mostly) bad of the technology, finished our dinners, and went on our separate ways.

But the question hung in the air with a slight modification: “How might I think about AI?” And that, in turn, took me to a phrase I use a lot: “Learn from the past, deal with the present, intercept the future.”

And so, here are some ponderings for Mark. The last pondering is ChatGPT musing about itself. You’re welcome to listen in.

Learn from the past

Let’s take a quick look at what we’ve learned from a range of past technologies. Here are a couple of thoughts.

The biggest consequences of a technology usually extend beyond its original purpose.

For example, the automobile was designed to improve transportation. But its invention resulted in suburbs, reshaped retail locations, land use, and daily routines.

AI might do analytical work well and cheaper. Will it also reduce the “apprenticeship” work that future leaders need?

Technologies eliminate tasks, create tasks, and redistribute work. But the balance is not automatic.

“Technology has always created jobs” is a popular but insufficient reassurance. A lot of factory workers lost jobs to earlier automation and the provided “computer programming” did not do them much good.

Ask our AI leaders, “Who captures the value, who bears the cost, and who has a voice in the decision?”

Deal with the present

AI is already here and is being used (or ordered) in many organizations. Here are a couple of thoughts.

Value is shifting along the workflow.

Look beyond individual tasks. If AI makes one activity faster or cheaper, where does the bottleneck move? Does verification become more important or is it being neglected?

Here’s a professional services question: If customers can obtain part of our expertise cheaply, what will they still value enough to purchase?

Review AI evidence on a regular schedule.

Here are four questions:

• What have we learned?
• What assumptions are slipping?
• What should we stop, change, or increase?
• What new signal deserves attention?

Intercept the future

The management sage, Peter Drucker, stated that “strategic planning deals with the futurity of present decisions.” That statement connects foresight to resource commitments and systematic feedback.

Again, here are a couple of comments.

Identify capabilities that cannot wait.

Some capabilities are transient and can be developed quickly through focused training. But some capabilities take years to build: judgement, deep expertise, managing organizational change, and trusted relationships are a few.

If you’ll need these capabilities in three years, when would you have to start building them.

Think about where internal and external customers will locate value.

Discuss what customers will need, expect, and pay for. Begin testing tomorrow’s value proposition while you can still finance the learning.

And so, I asked ChatGPT

I went to my intern and asked it to consider possible future outcomes from itself and its brethren. Here’s what it said:

“Possible futures: profound changes whose pathways remain uncertain

This is where you examine outcomes beyond the assumptions supporting your baseline:

• AI-assisted discovery dramatically expands humanity’s ability to address disease, energy constraints, and environmental damage.
• AI and robotics make human labor unnecessary for a large share of production, forcing societies to reconsider income, purpose, and participation.
• Dependence on automated systems erodes human competence and meaningful control over essential institutions.
• Highly capable AI, through misuse or loss of control, contributes to civilization-scale catastrophe.”

I really don’t like its fourth possibility.

In Greek mythology, Prometheus steals fire from the gods and gives it to humanity. Like Prometheus, we have been given fire.

The Promethean question for leaders is: As we acquire greater power to shape the world, are we developing the judgment to decide what world we should shape?

Are we?

Bill