The state of AI

As the principal of a small creative firm working at the intersection of design and technology, I think about AI a lot.

I recently watched an AI tool called “Midjourney” spin up a brand for a fictitious nonprofit using text-based prompts from a user. It took it (the AI tool) less than two minutes to design and the end result was quite good.

Needless to say, I was flooded with big questions about the future of our work.

I run a small team of project managers, developers, and designers and we deliver a product that on the surface, looks comparable to what AI just produced in the time it takes me to walk from my bedroom to the kitchen. Sure, the website produced by AI still required domain-specific knowledge to create the prompt, and “quality” design goes beyond aesthetics and involves strategy and art, but we are still in the early stages of this technology. In 10 years, after sufficient adoption and application of these models, I wouldn’t be surprised if humans never open an Adobe product again, unless, like a craftsman, they enjoy the experience.

If a domain such as design can be automated or outsourced to AI, I can only imagine the carnage that is coming to other domains that rely much more heavily on patterns to produce their outputs. Domains such as finance, which is highly data-driven and pattern-focused, or healthcare which involves very algorithmic diagnoses based on defined sets of variables, are ripe to be done by AI. They’ll be done faster, cheaper, and far more effectively than any human can.

A Hong-Kong gaming company named NetDragon Websoft replaced its CEO with an artificially intelligent algorithm that reviews business analytics, makes key strategic decisions, and handles risk management. The company has outcompeted the Hong Kong stock exchange and has seen its stock prices surge as a result.

The big question

The big question that everyone should be asking themselves right now is “can I be replaced by AI”? Tools like ChatGPT are being touted as major productivity boosters. They are presented as “assistants” that can help one person do the work it used to take 100 and proponents of AI are claiming that these tools will only serve to make our companies and economies more productive.

But what if you are one of those 99 other people that used to do that work?

Whose job is it to keep these people employed? If the CEO of a company has the option to replace 40% of its workforce with a virtually free AI solution, is it their responsibility to shareholders and customers to do so? As we’ve seen time and time again over the past 20 years, when company profits are pitted against social utility, profits usually win.

Or is it the responsibility of the workforce to make themselves irreplaceable? To modify a quote from Adam Smith:

“Man naturally desires not only to be employed but to be employable”.

As the owner of a company and the leader of people, the idea of replacing my team with AI seems distant and undesirable. However, I’d much rather have my team be irreplaceable than have to resist the convenience of replacing them. I think it is a safer bet for everyone, not just my team, to turn themselves into lynchpins rather than to count on the morality and virtue of owners, CEOs, and corporations.

So what does irreplaceability look like in the age of AI? If we are already seeing the replacement of accountants, journalists, truck drivers, physicians, lawyers, programmers, data analysts, and content creators, then what jobs are left? What is safe from the tidal wave of automation?

The cartographers and the artists.

Be the mapmaker

In the literal sense, I’m sure whatever cartographers have remained employed in the modern era will also be replaced by AI, however, I believe the metaphorical mapmaker will always be safe from replacement. At the most basic level, a cartographer’s mission is to travel into unknown territory and transform that unknown into something structured and usable by others. The cartographer goes to that place described by Shell Silverstein as “where the sidewalk ends”, and then keeps going.

AI operates on existing knowledge. These Large Language Models are fed a staggering amount of content from the internet, develop an ability to predict appropriate responses to prompts, and have their answers re-inforced or corrected by users. The key idea here is the existing knowledge that AI relies on to be useful.

In short, AI cannot innovate. It cannot make something out of nothing (at least not yet). It relies on a history of ideas in order to be useful, a map.

This means anyone who uses maps is in trouble. Administrators, order takers, number crunchers, process followers, and status quo people are all in trouble. Once the cartographer creates the map, AI won’t need anyone else. In the age of super sophisticated answer machines (AKA ChatGPT), does knowing something become obsolete? What is the value in your knowing when it takes me a few seconds to also “know”?

Today, the value lies in the question. This is where the cartographer excels. The cartographer is able to start with a big question and widdle it down to a strategy that can be applied by others (and increasingly AI).

Questions that start with “What if”, “Why”, and “How”.

To become comfortable with big questions is like becoming comfortable traveling down an unknown river or bushwhacking through a dense jungle. At first, the cartographer feels uncertain and alone. The nights in the jungle are long and the lack of companionship difficult. But eventually, as his competency increases, he starts to enjoy this process. The cartographer begins to crave this adventure. To need the unknown. It becomes the only way he feels useful. Alive.

Be an artist

There is a town in mainland China called Dafen Village that is responsible for 60% of the world’s paintings. Here, existing masterpieces are replicated by painters and sold to the mass marketplace. This Henry Ford-esque style of production involves individual painters focusing on a single element of the work and then passing the canvas to the next painter who specializes in the next compositional detail and so on and so on….

These are not artists. These are cogs in a machine.

These jobs are fodder for the AI productivity machine.

An artist is anyone who employs emotional labor to create something original. The artist is who creates the work, the painters in Dafen Village are just…laborers.

The artist is safe in the AI revolution because they employ emotional labor to create something original. An AI algorithm can’t do either of these things.

Which role do you play? Are you the painter or the artist?

If your work is downstream of the actual creation process, then you are a laborer. If you are a laborer who works on a computer, you’re in the most trouble. AI robots won’t be able to replace the DPW anytime soon, but digital task rabbits? They’ll be dinosaurs soon.

How much creativity do you exercise with your work? You know you’re in trouble if applying creativity to your work has a chance of getting you fired. How creative do you want your doctor to be? What about your accountant or the guy who installed the elevator in your building? Probably not too creative.

But creativity is the best form of job security. Your ability to connect two previously disconnected ideas into something novel and valuable is the human power that AI currently cannot defeat. But, it requires emotional labor. It requires sitting in front of a blank canvas for hours, weeks, months, years. It requires spending time on something only to throw it out and start again. It involves staring out your window in quiet contemplation, waiting for your mind to find a spark of something. It means going for a long walk in the woods. It means seeking inspiration.

Are we doomed?

AI seems to be growing in its applicability and workplace dominance right when our human workforce is at its weakest. Our collective attention and focus have been splintered into 20,000 pieces and most of us don’t dedicate enough mental capital to the work we do. Creativity requires struggle. It requires sustained effort over long periods of time. It’s experimentation and learning. It’s deep work.

But we live in a shallow age. Most of us don’t get more than 20 minutes of uninterrupted focus time per day. Even if we were allotted sufficient focus time, most of us have lost our ability to focus on one thing for more than 30 minutes. Our brains, the neuroplastic organs that they are, have long been rewired to crave constant hits of stimulus and novelty.

AI is going to crush most of us and our response as individuals and as a society is going to have to be developed in 15-minute bursts. We may be doomed.

Are you an automaton?

There are leading indicators for your future success in this new world we have entered. If you want to know what your prospects are, look at your bookshelf and your notebook. These are the inputs and outputs of the one thing AI does not have, your mind. The information and ideas you feed your brain are the raw material for your thinking and the writing in your notebook is the product you are uniquely capable of producing.

So the answer to irreplaceability in the age of AI is at once both remarkably simple and yet sufficiently hard.

Read and feed your mind.

Write and develop your ideas.