Let's talk about technology transitions
In the late 70's and early 80's, the introduction of personal computers started changing clerical work. People whose job it was to take handwritten drafts and turn them into documents discovered that work disappeared. Bookkeeping jobs also got hammered by spreadsheet software.
Many of those folks shifted to administrative assistant roles – managing schedules, coordinating projects and handling communications. The bookkeepers that adapted were the first folks we now think of as business analysts – using spreadsheets to model scenarios instead of recording transactions.
The same dynamic would appear again in the 80's and 90's for assembly line workers and machinists, especially in the automotive industry. Manufacturing automation and robotics changed the space dramatically.
The painful part of this shift was that many of those displaced workers never fully recovered economically. Some folks moved into logistics, or warehousing. Many left manufacturing completely – often for lower wages.
By the 2000's the internet and eCommerce would drive another major shift. Travel agents, stockbrokers, retail bank tellers and people who sold encyclopedias all saw their roles shrink or disappear. If your role was access to information, or the ability to process a transaction, the internet took it over.
Sure, some travel agents became travel advisors. Some stockbrokers became financial advisors. And some bank tellers transitioned into relationship banking. Unfortunately for those in the classified ad business – those jobs didn't transform, they just ended.
You can see the trend at play, right?
Also in the 2000's software ate the world. Desktop publishing killed the typesetter. Digital photography killed film processing (where Kodak went from 150,000 employees to bankruptcy). Digital recording collapsed the need for large recording studios and newspaper newsrooms were cut in half (at least).
The early adopters shifted to digital designers, Photoshop editors, independent producers, bloggers and podcasters. Some would say it was the drive of freelancer growth and the gig economy.
I don't have to continue, but I'll just add that I personally watched multiple of these transformations and sat in boardrooms when companies said they would never move their precious and private data off-premise to cloud platforms, but then AWS, Azure, and Google pushed right thru their objections.
The early adopters who saw what was coming shifted from sys admins to cloud architects and part of DevOps teams.
But here's the thing about those technology and automation shifts…
The speed of transition matters immensely. When the shift happens over a decade, people have time to adapt. When it happens in 2-3 years, the human cost is much higher because there isn't time for natural career evolution.
When domain knowledge remained valuable, even as task execution was automated, we saw transitions that we could say worked out. After all, in most cases the total economic value increased (even if the distribution was far less even).
When was the last time you looked at AI?
Maybe you tried AI in the form of ChatGPT a couple years ago. It hallucinated and you thought, “this will never replace me.” I remember feeling that way 3 years ago. It wasn't something I was using daily yet. I laughed at its mistakes, but kept playing.
And then one day the playing shifted to use, and then daily use. For real work.
And the scary thing is this…
Imagine if all those shifts I mentioned all happened at the same time. Not one every 5-10 years. But all at once. And not over a decade, but in a year or two. Imagine the consequence of all the jobs shifting – where career transition is so much harder because the transformation isn't in a specific industry but is happening in all of them, at the same time.
Honestly, I don't think people grasp how quickly things are changing.
A new technology transition is already underway
Three years ago I used ChatGPT to create different versions of a poem, in different voices, and it was fun. But it was a toy.
Today I worked with a colleague to create a new piece of software. We had worked on the rules, guidelines, and algorithms. A total of 10 files with data, rules, and structures. A way to describe what we wanted to do – over tons of data. We loaded it up into Claude Code and had it review the files.
We didn't have to explain much except where we wanted this new web application to be hosted, using what kind of technology. Then it spun up 5 sub-agents and built the software. When I asked it to test itself (comparing its output to a finished sample we'd imagined), it found 7 bugs and fixed them. Then it started testing itself further, and fixed a few more.
Then I told it I didn't like how the page breaks of the PDF that was being output felt off. So it started reviewing the output visually and dynamically changing things until it felt like it was worth me looking at.
And then it was done.
The whole project took less than 4 hours. And the code is good. Solid. Not needing any advice from me.
Is the future all doom and gloom?
Am I just raining on your parade? Predicting gobal poverty and the rise of the machines? No. Not at all.
But I am suggesting you dig in and figure out this AI thing as early and as fast as you can.
Because the early adopters are the only ones that have ever managed to make it in these transitions.
Four ways to embrace this transition
First, find the things that you spend a lot of time on. You'll likely know those best, and be best able to determine how close AI is to helping you with them (or how far out it is). But don't give it the silly stuff. Dig in and really challenge it – you'll be surprised.
Second, pay for a premium account. It's going to cost you $20/month but if you're only playing with the free version of Claude or ChatGPT or Gemini, you're not experiencing what's really going on.
Third, make sure that you never stop investing in relationships. Your people skills will never be more valuable and more useful than as we navigate this transition. People trust people. And remember that in some of those transitions, it was the relational and community-oriented roles that survived.
Fourth, find an AI friend. Note I didn't say make an AI friend. I'm not suggesting you build an AI avatar to be your friend. I'm talking a flesh and blood person.
You've heard people say it's not worth it to buy a boat, just get a friend who already has one. The same is true with AI. Find someone ahead of you on the journey and take them to lunch. Often. It will accelerate your journey and help you skip the potholes.