Summary

AI is coming for many jobs. Bullshit jobs will be the first to go. Those of us with a real job must cultivate a few characteristics to be AI-first in the way we work and stay relevant in the job market.

In a previous blog post, I noted how I’m seeing the shape of teams changing, with AI. In today’s post, I want to offer a short list of characteristics that I consider essential for the AI-first knowledge worker. But first, a few caveats.

Are these characteristics exhaustive?

No! I might write a follow-up soon.

Do these characteristics guarantee success?

No! Success in corporate careers is the product of a multivariate equation. The individual is only one variable in that equation.

Will cultivating these characteristics protect you from redundancy?

Probably, but not certainly. You could be in the wrong part of an organisation at the wrong time and lose your job, despite being highly skilled.

Since I’m not a soothsayer, how seriously should you take my list?
As seriously as you’d take a random guy on the internet. We’re all allowed our opinions, and this is mine.

With those disclaimers out of the way, let’s get into my list. If I’ve piqued your interest, I urge you to ask how many of the characteristics apply to you.

0. You have a real job

Before getting into any of the other characteristics, let’s be clear. Unless you back yourself to handwave about AI, you need a real job. Not a bullshit job. Aravind Srinivas is already talking about two job categories that’ll disappear in six months. The last thing you want is to be in a non-job job.

Do what you can to find yourself a real job, where you get to build stuff and deliver value with AI. Even if a bullshit job comes with a promotion, avoid it if you can. When job cuts come, they’ll come for the bullshit jobs first.

1. You’ve cultivated a maker’s mindset

Code wins arguments. Prototypes spark imagination. AI makes it easier than ever to build a portfolio of work that represents you. To harness that power, you must first think of yourself as a maker. Even if your outputs don’t always make it to prime time, consider how your eyes on a screen and fingers on a keyboard can create stuff. You’ll drive more value that way, even if your job is to “lead” other people. So stop fluffing and start building.

2. You work async-first

In his book, Co-intelligence, Ethan Mollick has the following chapters.

AI as a person
AI as a creative
AI as a coworker
AI as a tutor
AI as a coach

You can simulate an entire team of collaborators with AI, provided you are willing to be async-first, and you’ll make far more progress alone than if you jump on Zoom calls at every opportunity. So, stop wasting your time and everyone else’s. For example, in the short video below, I show you how you can use two AI agents - one for UI design and another for coding - to build a prototype. Think about your work and how you can progress asynchronously, whenever possible.

3. You can judge when to synchronise

While you can go fast alone, you can go far together. Synchronisation is inevitable when you work in a company. We’re emotional and political creatures. AI still can’t help you deal with those aspects of working with other human beings. So, develop judgment to decide when you must synchronise with your colleagues. Async-first is still not async-only, especially when you work in a corporation.

Effective collaboration balances asynchronous and synchronous ways of work

4. You can delegate effectively

Talking to a chatbot is the most ad-hoc way to use AI. I do it dozens of times each day, but get most value when I can hand off a seemingly complex job to an AI agent. For example, the other day, I asked Perplexity Comet to study the latest report by the Kalahari Leopard project and to scrape information from their Facebook page, to tell me where I could look for the well-known leopards of the Kgalagadi Transfrontier Park. It’s an activity that could take hours, but through thoughtful delegation, I had Perplexity do the job while I watched a movie with my family.

Think about task breakdown. Think about repeatable activities that you can automate through shortcuts, AI automation, and agents. For that, you’ll first have to be a champion at delegating work.

5. You’re a skilful business writer

Writing is still the primary way to instruct AI with precision. You can’t delegate well if you can’t write clearly. If you scoffed at me when I first said that writing is a superpower, take me seriously now. Writing is the new programming language. Specifications trump prompts and code. To quote Sean Grove, of OpenAI:

“A written specification aligns humans and machines. Without it, you just have a vague idea.”

And hey, the idea is to write clearly. AI can help you if you know the fundamentals.

6. You’re a critical thinker

If you’re a skilful AI user, your employers will probably ask you to solve business and team problems with AI. Since AI is “so easy”, it’ll be tempting to slap AI onto every broken process you see around you. As Sidu Ponappa says, with AI, we’re now fighting “a battle against entropy”. You’re faster, but you also hit the wall faster. Problems that would take months and years to surface now emerge in hours, days and weeks.

As a critical thinker, you expect these problems and build guardrails against them. Whether it’s by placing humans in the loop or reimagining processes from the ground up, or through designing thoughtful AI workflows, the idea is always to be ready for the battle against entropy.

7. You see through the hype

Technologists love new tools and tech. And once you have a new tech hammer, it’s tempting to go looking for nails. It’s ok to look for nails, but as Mathew Granade of Dragonfly Insurance says, be “problem-driven”, not “technology-driven”. Every new feature from an AI vendor won’t be meaningful for your team or business. You can’t always be experimenting either. Look beyond the marketing spiels and focus on value instead.

8. You embrace perpetual betas

AI is imperfect. Hallucination is a feature, not a bug. When hallucinations work in our favour, we call it “intelligence”. If you can’t take AI imperfections in your stride, AI will forever frustrate you.

Don’t compare your best outputs with an AI’s worst outputs. Instead, think about the tedious, predictable, easy-to-define work that AI take off your plate. Don’t aim for perfection. Instead, build safeguards for errors, so you reduce the cost of AI boo-boos. Think about AI as a hardworking, bright intern who will, on most days, take mundane work off your plate and give you time back for creative work. When it fails, as it inevitably will, cut it some slack.

9. You have taste

AI gives you infinite options, but it takes human taste, curation, judgment, and discernment to identify what is truly valuable. Taste is now as important as technical skill. Here’s what Sari Azout says about the topic.

“AI is powerful but taste-blind. It can make anything, but it has no idea what's worth making. But AI + your taste? That's the game-changer. The more AI can execute, the more your eye for what's interesting—your ability to discern and curate what matters and why—becomes everything.”

Enough said, eh?

10. You’re forever curious

I don’t remember a tech trend that’s moved faster than the current AI revolution. We discover new possibilities overnight. The models are becoming more capable with each iteration, and as a result, AI-powered products and agents are delivering new features and tools at a breakneck speed. Of course, we must resist AI FOMO, but at the same time, retain the tinkering mindset. The more curious we are, the more likely we are to uncover innovative ways to achieve outcomes with AI.

So there you have it - my current list of 10 characteristics of the AI-first knowledge worker. I’ll continue to revisit this list over the next several weeks and months, but for now, I’d love to seek your agreement, disagreement and comments about this list. Let me know what you think.

And here are a few parting words. AI is rewriting the rules, but it needn’t rewrite your critical thinking. Tools are fleeting, but the fundamentals endure. Stay sharp. Stay sceptical. Free up your calendars. Free up your mental space. Build, write, question. And, most importantly, never outsource your taste or your curiosity. In the age of AI, that’s your real edge.

10 characteristics of the AI-first knowledge worker