Staying 'relevant' in the age of AI

First published:

careers AI

There's increasing concern about the impact that AI is going to have on the job market, especially on ‘white-collar’ jobs. Try searching “AI” and “white-collar” on Google News to see what I mean: you’ll find plenty of doom-laden headlines.

Why we should panic

I share this concern. In a previous blog post I despaired that, “any task that can be done online is likely to be one that AI will eventually be better at than humans.” Recently I listened to Geoffrey Hinton, Nobel Laureate and the "Godfather of AI” echo this pessimism. He told Stephen Bartlett on his podcast that, "a good bet would be to be a plumber.”

If "learn to code" was the career advice of the 2010s, "learn to plumb" might well become the career trope of this decade.

"Learn to code" seemed like wise advice 10 years ago. I have a friend for whom it was. He took the gamble of leaving his old job, invested in retraining at a coding bootcamp and got in at the right time. The outcome today might be very different. There is all round panic about the impact of AI on entry-level coding and other jobs. Having recently tried out some AI enabled code editors such as Cursor, I can see why.

AI is already replacing or reducing many tasks that people, including myself, used to do for a living. For example, in my own case: internal comms. I used to joke that internal comms was "writing the all-staff emails that nobody else wanted to write". Now Gemini, Claude, CoPilot and ChatGPT can all help make us internal comms specialists. The same applies for marketing and a lot of other 'comms' roles. They might not disappear entirely, but there will be less of them.

Why we shouldn’t panic

It may feel as if the pace of AI's progress is unstoppable but the known laws of physics suggest that this is not going to be the case. In a world of limited resources, even exponential growth comes to an end.

I don't know exactly when or how the slowdown will appear, but it could arise from one or more of the following scenarios:

  • Increased cost of electricity, computer hardware and access to natural resources like water
  • Decreasing quantity and quality of training data for future systems. Large language models depend on content. If the people who create this data (aka humans) give up because there is no longer a market or audience for their work, what then?
  • Inherent technical limitations to the existing data and hardware stack
  • Climate / environmental / geopolitical / economic instability or other "black swan" events. (To be fair, if these are the reasons, we’ll have other problems than AI taking our jobs!)

For the avoidance of doubt, I'm saying there will be a slowdown. An equilibrium, not a shutdown. We won't go back to the way things were before. By the time things calm down, some of the things that used to count as “work” will be gone.

When the AI equilibrium happens, we will be better placed to take stock of the things it can't do as well as humans. And I’m pretty sure that there will be plenty of areas of intellectual activity where humans can still add real value.

Until then, it's a matter of holding our nerve, optimising our options, and not allowing ourselves to become de-skilled. Coders need to keep coding. Thinkers need to keep thinking. Writers need to keep writing.

One of the reasons I am writing this blog post the old-fashioned way is that I don't want to lose my ability, albeit modest, to express myself in writing.

What new skills will we need?

It's not just about holding on to existing skills; we need to be committed to developing new ones.

The answer to “what new skills do I need to learn to remain relevant in the age of AI” is going to be different for everyone. So I don't think the talk about "learning to plumb" or pursuing other types of 'blue-collar' work is particularly helpful. It will lead to bandwagons and I don't think it's helpful talking about white-collar' or 'blue-collar work' any more. We're not living in 1950s America. (Note to self: stop talking about 'blue-collar' and 'white-collar' work from now on :-)

Here's the thing: manual work is the original digital work. It's Digital 1.0. Even our word ‘digital’ comes from the Latin word for ‘finger’. What if the "killer app" that gave homo sapiens our edge is not so much our brain size but our dexterity?

If this is the case then the space where knowledge work breaks away from the computer screen and into the real world is the one to be in. What spaces do our own interests and aptitudes lead us into?

That’s my current line of enquiry.

I’m thinking aloud via these blog posts so I don’t have any answers yet. In fact, I'm still working out what I think about AI. I may swing from doom to optimism to defiance from post to post, even paragraph to paragraph and I’m ok with that for now.

Medium-to-longer-term strategies for staying relevant

What I can offer at the moment is a few hunches for strategies that might help to keep people like me relevant for the medium-to-longer term.

  • Focus on essential industries - food, water, energy. These are not going anywhere. What knowledge work, especially augmented by AI, can add value to these sectors? For example, identifying and capturing new sources of data for AI to work on. This requires human planning and input, as well as hands-on experience in the field.
  • Become expert in information and AI governance, privacy, cybersecurity, ISO standards etc. The internet is likely to change and become a lot more regulated. *Experts who keep abreast of these changes and offer practical solutions to navigate them will be in demand.
  • Upskill in on-premises. I’m starting to think that the “software as a service” trend has peaked. Some organisations, especially those with valuable intellectual property, are going to be increasingly cagey about entrusting their data to someone else’s cloud. Even the use of AI may be taken ‘in-house’. I might be wrong about this and it would turn upside down the trend of the last decade or so. A lot of corporate IT departments have "shut down the metal" and outsourced their infrastructure to third party cloud providers. Should there ever be an about-turn, even a partial one, then old-school server management and networking skills will be back in fashion. What new skills and innovations could we bring to such scenarios?

Short-to-medium term strategies for staying relevant

For those of us who need to focus on a shorter time window (short-to-medium term, or just short), here’s what I am currently considering:

  • Don’t give up / give up on the day job (yet). It may be harder out there to find work at the moment but that’s more to do with all too human factors in the economy than the AI bot waiting to take your job. Your old roles are not going to disappear overnight and your next role might look quite a lot like your last role. A major pivot might not yet be necessary: but can we use this time to skill up in something new and maybe get AI to help us?
  • Be willing to consider geography again. The era of remote working and digital nomadism has peaked. In the age of AI, the kind of roles that can be done 100% remotely are as vulnerable to automation as they once were to being off-shored. (I've benefited from remote working opportunities over the last few years, so I don’t say this gladly). Someone once said that the future belongs to those who turn up for it and being able and willing to “turn up” is going to be crucial in coming years. Sadly this means that the distribution of opportunities will not be evenly spaced out. Not everywhere can be Palo Alto. We humans are creatures of our geography. How can we make our geography a strength and not a weakness, and is there anything AI can do to help?
  • Stand on the shoulders of giants. I recently read that the AI company Anthropic had cut up and scanned millions of books to train its large language model Claude. In my own house there are maybe about 200 books and I haven’t read all of them, which is a source of frustration. So I am thrilled that I can converse with such an erudite system that has absorbed so much human knowledge and wisdom in a way that I never can. Large language models are standing on the shoulders of giants - just like us. Whatever new challenges that AI will bring, it’s also a wonderful and unprecedented opportunity for those of us who want it. What new things might we glimpse from the shoulders of those giants?
  • Get certified. That rolled up university degree at the back of the sock drawer hasn’t seen a lot of action recently, has it? Rather than focus on degrees (unless our chosen profession really requires it), many of us might be better off looking at micro-credentials and certifications and building a bespoke portfolio education for ourselves. I will explore this in more depth in a future post.

So this is where I am currently at. There’s a few micro-credentials on my to-do list now so I’d better get back to work 🙂

Thanks for reading!

TL;DR

"Like many, Graham has concerns about AI's impact on 'brain work,' especially in white-collar jobs. Yet, he argues that complete panic isn't necessary; real-world resource limits will likely temper AI's exponential growth. For Graham, staying relevant means focusing on essential industries (like food, water, energy) where human dexterity and on-the-ground knowledge are invaluable, even when augmented by AI. He believes that 'on-site' rather than remote roles are less vulnerable to being replaced by AI and advocates for continuous micro-credentialing over traditional degrees. Ultimately, it’s about adapting, learning, and finding genuine, hands-on ways to contribute in a world rapidly reshaped by AI."

Thanks to Google Gemini for the TL;DR version!

Last updated: