What Are We Working For?

“Gradually, and then suddenly,” is the rueful but accurate way that a certain Hemingway character describes how he went bankrupt. It’s also an apt description of how artificial intelligence came to be so central to our work. And not just to our work, but to our attitudes about work, about what we are meant to do and what we expect to gain from the effort. The comparison with bankruptcy is cautionary - but maybe more than that.

Like money, the danger of AI lies in the choices it causes us to make. Is it an asset to be accessed judiciously - like Gilded Age industrial barons? Or is it a concept, to be exploited strategically like arbitrageurs? There are no simple answers. AI is not just the evolving and expanding volume of data: it is the result of functioning neural networks and large language models, and the availability and velocity of supercomputing.

It's impossible to deny the value of artificial intelligence to the world we inhabit now. Most people’s awareness came by way of the retailers using our data trails to pitch purchasing ideas, compiling records of what we’ve bought and even what we considered buying.

By the time our data became useful to Amazon or Target we had already been codified by banks and credit agencies. There are parallel versions of each of us coming to life in databases all over the world, existing only to help someone anticipate what we decisions we’re facing or soon will face, and to find some way to influence our choices.

The advantage to you or me in all this is that our choice may be simpler of more satisfying. And the retailer evaluating the sale amid dozens or millions of other sales can learn what the market values, what it might prefer to current choices, and how to price things. And our influencers not only gain the sales, but more familiarity, more insight to our thought processes - a firmer point of contact. It’s probably pointless to wonder who gains the most.  

AI’s value in scientific and medical research, and of course in manufacturing, is clearer. The accumulation of knowledge about energy and climate; about structural engineering, material science; and particularly in medical diagnostics and pharmaceutics are all positive. AI makes scientific progress faster and offers lifesaving possibilities.

AI in manufacturing needs no defense either. Data analysis increases industrial productivity by improving predictive maintenance, supply chain management, plant and worker safety, and identifying wasted resources and energy.

In science and technology what AI is improving is human understanding - though that is the glass-half-full way of saying AI is removing human inefficiencies, human limitations, and human costs. The threat that automation poses to laborers is equivalent to AI’s threat to engineers, designers, programmers, financial analysts and planners, and dozens of other positions that have lately come to known as “knowledge workers.” They have that title now - but who needs them if the function and cost can be defined more optimally in favor of an automated tool.

And even the most ordinary but necessary tasks can be assigned to an AI tool: Write my homework essay. … Draft my resignation message. But is the speed or efficiency of a task the reason for completing it? Does the gradual improvement of a person’s math skills or language proficiency have no value?

The interruption of human development is not a forecast possibility. It is a work in progress. The threat of AI is not simply to the work we do but to our place in a universe that has become a system, where outcomes are determined automatically according to past examples. Our influence on those outcomes is diminished and diminishing, leaving us subordinate to the system, which offers no appeal.

At the same time AI makes socialization less and less necessary, draining away the necessity to learn from others, to share information, to help others learn, and to develop skills together.

And what human qualities are we likely to lose in such a system? Ethical judgment and moral authority seem to have diminishing value now, and there’s no sign they will feature in an AI-guided world. It can define and impose order, but how can it know when to overlook the rules?

Creativity and artistry will become rarer than ever, because AI works with established results and patterns. Do not expect it to alter its direction.

And AI should not be counted on for invention or counter-intuitive proposals. It is ordered toward organization, not innovation.

Judgment, creativity, and adaptive reasoning are qualities that rarely occur in together in single being. They emerge much more frequently in organizations, among groups that learn to work together and build trust among themselves - in each other and in what they accomplish. AI suggests that partnerships, families, and teams, etc., are fruitless approaches to living. Considering the speed of AI’s progress in the past decade there’s reason to believe that people do not object to this.

We have come to this point in human development rather obviously, and we may be only gradually recognizing the power we have set loose. But suddenly, we will.