Are you above or below the AI? Your job may depend on it.

This old time photo of blacksmiths shows that Just as industrialization automated work done by blacksmiths, AI will automate many of the tasks done by today's information age workers. (Image credit: Randy McPherson/Pinterest)

Just as industrialization automated work done by blacksmiths, AI will automate many of the tasks done by today's information age workers. (Image credit: Randy McPherson/Pinterest)

We see it everywhere: it’s a headline like “Automation threatens 800 Million jobs,” it’s a Quora topic full of accusatory and worried questions, it’s a Wikipedia page dedicated to Technical Unemployment. Everywhere we turn, technology is expanding into new industries and functions, and it seems like wherever technology goes, markets and jobs disappear. So what exactly is going on?

Let’s focus on two terms that are often at the core of this discussion: artificial intelligence (AI) and application programming interfaces (APIs). The former allows computers to make decisions based on unfathomable amounts of data and pre-trained programming that looks for certain clues and statistical outliers, while the latter allows computers to execute decisions from anywhere in the world to any internet-enabled device. Combined, they represent the brain and the muscle of many modern innovations and at the same time, the underbelly of many complaints.

Do certain jobs exist today that will no longer exist (or will be relegated to a hobbyist pleasure) in the next 10, 20 or 30 years? Of course. On my walk to work this morning, I didn’t see a single blacksmith, leathersmith or horse stable, despite the fact that I live in one of Paris’ historic districts. I work for a startup based in industrial space formerly used for garment manufacturing. Just as the industrial age phased out the need for blacksmiths with the advent of the factory line, the artificial intelligence age is phasing out many jobs once thought to be secure in the information age: lawyers, accountants, even journalists. (Perhaps many of these functions are actually artifacts of the industrial age.)

One of the golden rules of jobs “disappearing” at the turn of the 20th century was that if 1,000 or more of something could be bought in a day, someone would find a way to make those somethings at scale, more cheaply than an artisan could with similar quality of materials.

So what’s the golden rule for job displacement today, when AI threatens to replace jobs across so many industries? And above all, how will you be affected? Your fate depends on the answers to questions about your priorities, job role, industry, and your willingness to take on new skills and responsibilities.

Who (or what) determines your priorities?

Think about the objectives you have in your work right now, the metrics used to quantify your performance and success. Next, think about the means by which you acquire the knowledge or the ability to achieve success. What’s enabling you to work? What’s just before you on the workflow chain? Are you being fed data by a machine and reacting to it, or are you being fed problems by a superior and being told to solve them?

If what determines your priorities is someone feeding you a “why” problem and letting you determine “how” to solve it, then you’re above the AI. If something is feeding you a “what” problem — a support ticket routed to you, an alert to a poorly performing unit, a new sales prospect to follow-up with — and you’re determining “how” to react, then you may very well be training data for the next generation of AI.

AI is fantastic at processing “what” problems, prioritizing them and determining the best solutions. The quality of these solutions can be quantified at the end: customer satisfaction and retention, revenue gained, meetings booked. If your job is to qualify leads, you may be replaced by a predictive lead scoring solution that will eat all your historical data and look for statistical correlations between firmographic profile data and success, where “success” is whatever end-step you want to look at: getting on the phone with an account executive to build a lasting 1:1 relationship, inputting a credit card, or trying a new feature.

If a machine is determining your priorities, it’s only a matter of time before it learns to prioritize and solve them instead of you.

Are you serving or creating?

Another great question to ask yourself is whether the function you perform is to create a product or whether it is to provide a service. Humans create products, but AI productizes a service.

A good example is language translation. AI is exceedingly good at productizing the service of translating, say, a menu or a phrase. However, human translators continue to create products like a French-language script for an English-language TV show that a French voice-over actor will then read to make the final product feel as human as possible. When AIs try to create, we get an uncanny valley feeling that something is off with the product, but we’re not quite sure what that something is.

Google Translate can process image data, run it through servers and feedback images that match the original text to produce a real-time translation on your phone using just your camera, sure. But Google has yet to create an adaption of a foreign film or TV show because that process is still a creative-driven process.

Where else do we see this?

In the industrial sector, there are people who tell machines what to do — programmers, workflow designers, and automation experts — and there are people who are doing what a machine tells them to do — assembly line workers, truck loaders, and deliverers. As AI and APIs continue to invade every stage of industrial operations, automation experts will expand their responsibilities, while the assembly line and logistics elements will continue to get automated step-by-step.

In retail and ecommerce, there are those who create assets for marketing campaigns, and there are those who are adjusting the deployment of those assets based on analytics supplied by, you guessed it, an API. Those analytics will continue to get more sophisticated, supplanting the need for someone to interpret them and creating an even greater need for asset creation as AI enables relevant conversations at scale.

Are you driving someone (taxi) or something (long-haul) to somewhere based on what Google Maps tells you? That’s bad news if your livelihood depends on it. Some variant of Google Maps will deliver those goods and people for you someday soon using Waymo.

Even in the software industry itself, if you’re being routed leads as a sales development person, qualifying them and pushing them to an email sequence and waiting for responses, your job tomorrow is going to look a lot more like a call center than a sales center. If you’re not solving complex sales problems by building lasting 1:1 relationships, you’ll be automated away soon.

Staying above sea level

AIs and APIs are like a rising sea level: everyone above them will continue to rise and expand their roles, while everyone below them will have to either swim upwards or drown. When factories were first opened and smelting machines needed to be managed, factory owners hired blacksmiths — people with smelting experience, and had them adapt their sector expertise to new tools.

This is why one of the biggest challenges today is to rapidly retrain the world’s workforce within the next several years. One such entrepreneur in France working on that is OpenClassroom’s cofounder and CEO Pierre Dubuc, who points out that in the short- and medium-term, more than 10% of the workforce will need to be retrained or learn new skills for new jobs altogether.

One way to float above this rising sea level is to take on more complex challenges. In short, you need to stay ahead of AI.

AI can drive on freeways, sometimes, but it can’t drive in complex environments — yet. Disney can train a robot to perform stunts, but these robots can’t design and coordinate stunts and pyrotechnics to make audiences believe the stunt. If you’re a truck driver on crowded streets or a Hollywood stunt producer, then you’re out in front, for now, and you’ll need to find a way to stay there.

The world is always eliminating its needs for certain skills, either because innovation shifts the focus to a new medium, or because the amount of time required to execute a task no longer requires a full-time employee. However, that doesn’t mean our growing population with all of its crises and macroeconomic problems doesn’t need humans to address them with creative solutions.

Find a problem that’s truly worthy of the human brain and you’ll find a job for life — or at least until a machine automates that solution.