Who gets to be an AI expert? Part 3

Who are the real AI experts, and how do we identify them?

I grew up surrounded by different kinds of expertise.

I lived near three research universities. These schools attracted the best and brightest from all over the world, and it was rare for me to meet anyone without at least an undergraduate degree.

In some ways, my dad was like many people who lived in my hometown. He was an MD and worked late nights in the emergency room. As a kid, I would sneak downstairs late at night and find my dad in the kitchen, eating “breakfast.” He was always reading a book, killing time while his body resolved its day-night confusion.

I don’t remember him ever greeting me. He would look up, notice I was there, and start telling me about his book. I would get a snack and settle in to listen to the night’s lecture. It was our ritual for many years.

Looking back, I’m amazed at how much my dad learned from his books. He read history, economics, and geopolitics. He spent ten years reading American history books and knew more about it than my teachers. When my brother switched his major to Economics, my dad tutored him despite having never taken an economics class in his life.

Technically, my dad’s academic background was in chemistry. I found his Masters thesis on a shelf one day.

“Oh, I don’t understand any of that stuff anymore,” he said when I asked him about it. He could answer any question about American history, but chemistry? Not anymore.

My dad was something of a paradox. On the one hand, he was an MD, a highly qualified professional. On the other hand, he was self-taught in fields outside of what he studied. In other words, he embodied different kinds of expertise.

How do we tell the real experts?

I learned important lessons from my dad. One was not to be blinded by the letters after a person’s name. Another was how much a dedicated person can teach themselves. He taught me expertise was more complicated than I thought.

I’ve written two articles about expertise and who gets to call themselves an AI expert. In the first, I called out the behavior of many so-called AI experts who appeared overnight after ChatGPT was released. In the second, I talked about the aesthetics of expertise and how some people get shut out of the conversation because they don’t look like experts.

In this article, I want to talk about genuine expertise. Who are the real AI experts, and how do we identify them?

The academic expert

This type of expert is easy to identify — they have an advanced degree in the field they claim expertise in. Living near three universities, I know plenty of academic experts.

In my last article, I wrote about meeting Engineering school faculty. These professors had decades of teaching and research experience. They were currently wrestling with a big problem in the tech industry: how to make deep learning models more efficient.

There was no doubt in my mind that these professors were experts. They’d spent years mastering their field and had a laundry list of publications to prove it.

Sometimes, academic experts don’t stay in academia. My older brother earned his Master’s in Computer Science before jumping straight into industry. He’d gotten an advanced degree but never wanted to become a professor.

Many graduates in fields like Computer Science, Data Science, and Information Science make this choice. Academia can be competitive, underpaid, and soul-crushingly difficult. Why stay when they could make good money working for a tech company? This type of expert overlaps with the “industry expert,” which I’ll discuss next.

Beware the wolf in tweed clothing.

Just because someone has an advanced degree in one subject doesn’t mean they can claim expertise in another. This should be obvious, but people do it all the time.

Yuval Noah Harari is the best-selling author of the book Sapiens. His background is in history and philosophy. So, why is he giving public opinions about AI? What authority does he have to speak on that subject?

I see this kind of behavior all the time from public intellectuals. They’re used to being asked what they think, so they’re comfortable giving their opinions on all sorts of things, regardless of their qualifications.

Just because someone is a doctor doesn’t mean they’re an expert. What exactly is their doctorate in? And have they done any meaningful work in their field since? Remember my dad, who can’t remember anything about his chemistry degree because he hasn’t used it in thirty years. Other academics have a solid theoretical background but little experience applying their theories in the real world. Take what they say with a grain of salt, too.

If you’re not sure about someone’s background, look them up. If they’re legitimate academics, you should be able to find their degrees and publications fairly easily. If not, be careful. They might not know as much as they claim.

A caveat: a person doesn’t necessarily need an advanced degree to be an expert. Some people disagree and think that only PhD’s have the right to call themselves authorities. Usually, I hear this opinion from people with many letters after their names (lots of “alphabet soup,” as my friend calls it).

I’ve never believed this. How could I, growing up with my dad? If academic credentials formed the only valid form of expertise, then I could end this article right here. But there are other legitimate experts out there who are worth listening to.

We’ll talk about the industry expert next.

The industry expert

Okay, if someone has an advanced degree, they’re an expert (at least, in most cases). Easy enough. But what about all those people working in the tech industry? Aren’t they experts, too?

Yes, many of them are experts. There are also plenty of people like my brother who work corporate jobs after earing their degrees, so there’s overlap between these groups.

The industry expert might have a computer science degree, or they might have slid into their career sideways. Some might even be self-taught. Either way, they’ve spent years working for tech companies solving real-world problems.

This kind of expert is sometimes easy to identify. Oh, someone works at OpenAI? At Google? For an AI startup? They probably know what they’re talking about.

My husband is a great example of this kind of expert. He worked in the tech industry for fifteen years before retiring to focus on AI research and teaching full-time. His hands-on experience in data centers and training AI models informs the opinions he shares with the world.

Just vibes

Not all industry experts are created equal, however. Some tech people work for innovative companies actively research and build new AI tools. Others experiment on their own or with their teams, trying to integrate AI into their workflows. Still others don’t work with AI at all. It all depends on where and how they work.

If someone works in the tech industry, but neither they nor their company uses AI, they might not know much about it. Sure, a general tech background is a useful starting point for understanding AI, but that’s all. Remember that this technology is still new, and even most tech workers only heard about it two years ago. They haven’t had any longer than you to learn about it.

And then there are people who deliberately distort their experience. In my last article, I talked about the tech bro aesthetic. Some people try to dress, talk, and act like tech insiders to project authority they haven’t earned. It’s just a vibe.

So, be careful which industry experts you listen to. Ask yourself what a person’s background is, including both their schooling and work experience. Have they worked in a data center? With AI tools? For an innovative company? With LinkedIn, this should all be easy enough to find.

Many tech professionals participate in open-source projects or publish their portfolios online. These are also good places to look for proof of someone’s expertise. Many self-taught industry experts rely on this kind of evidence to demonstrate their skills.

The adjacent expert

I want to talk about one more type of AI expert that I see a lot of these days — the adjacent expert.

The adjacent expert is not actually an AI expert. They are an expert in another field, and they comment on how AI will affect their industry. For example, my husband knows a woman who’s spent her career working in product development in Silicon Valley. When ChatGPT came out, she learned everything she could about it so she could understand how AI would impact product development. Two years of hard study later, she is qualified to discuss AI from a product perspective.

This kind of expertise is very common. A practicing lawyer gets asked how AI will affect law. A high school teacher comments on how her students are using AI. A philosopher shares an opinion about AI and sentience.

These perspectives are extremely valuable. After all, even the most informed Silicon Valley insider can’t predict how AI will shape every industry.

This is also a big part of the value experts offer. How will a societal change affect you? Your industry? Your specific corner of the world? This is where experienced professionals with different backgrounds can weigh in.

Crossing the line

Adjacent expertise is fine until it suddenly isn’t. Just like academics who step outside their PhD’s, sometimes adjacent experts act like they know more about AI than they actually do.

Many of the so-called AI experts I identified in my first article are adjacent experts who crossed this line. Marketing professionals who start selling courses on AI. Entrepreneurs trying to find investment for their startup. Grifters who promise to teach you their six-figure method for making money with AI. These people may not know much about AI at all. They’re just trying to sell you something.

Pay attention to how adjacent experts talk. Do they present themselves as professionals who are exploring changes in their field, or do they pass themselves off as true AI experts? Think about their motives, too. Are they trying to raise awareness or pass regulations for their industry, or are they trying to sell you snake oil?

Becoming an expert

These are the three most common types of experts I see weighing in on AI. Maybe you recognize some of them. Maybe you are one of them.

And maybe you want to be.

Many people want to learn about AI right now. They want to stay relevant in their current jobs, explore an exciting new technology, or help shape the future. If you’re starting from scratch, how can you become an AI expert? A real one?

You can go back to school to become an academic expert. Depending on your background and finances, you might be able to switch careers to develop industry expertise. For many people, however, these are not realistic options.

I recommend developing your adjacent expertise. Consider your education and experience, and think about how AI will change your field. Learn as much as you can about AI in your current context. Start to form opinions, but make sure you can back them up. Start to experiment and work on projects on your own. As you gain confidence, don’t rest on your laurels. Keep learning and keep growing.

Whatever you do, do not claim to know more than you actually do. That’s the fastest way to kill your credibility.

Expertise takes time. It’s tempting to rush to master AI as fast as possible, to fire off half-formed opinions, and to present yourself as an expert before you’re ready. It may feel like you’re racing the clock, trying to keep ahead of this new technology before it’s too late.

Don’t race AI. Instead, grow with it, starting wherever you are today.

At the end of the day, expertise isn’t something you can rush. Time well-spent, knowledge accumulated over the years, experiences remembered and learned from — these add up to real wisdom. In the end, wisdom is why you’ll stand out from the crowd of fake experts. You will have something truly meaningful to say.

I want to end this article where I began: with that image of my dad at the kitchen table. It’s amazing what a passionate, dedicated person can teach themselves. By the time I’m my dad’s age, I hope I know half of what he does. The time will pass, and the future we’re looking forward to or dreading will come. We might as well spend the time learning something worthwhile.

---

Read the whole series - “Who gets to be an AI expert?”

Part 1: Who do we listen to, and why does it matter?

Part 2: Do I look like an expert to you?