AI is the tech that everyone’s talking about these days. Visionaries are saying AI will change everything. Okay, that’s cool, but just a second. Will AI change your job? Will AI replace your job? What about your children’s future? Is AI coming for you?
As software engineers, we’re the ones on the cutting edge of AI development. Previous software innovations like automation, the internet, smartphones, and streaming changed our society and created great wealth for the software professionals who produced them. But the cutting edge of AI is different. It cuts both ways.
AI already makes writing and manipulating code faster and easier. What does that mean for the job of a software engineer today? What about for our children? What about me? I still write this column unassisted (no AI). Will AI terminate me? Let’s talk about it, person to person.
Eric Aside
For more on dealing with a constantly changing world, read Proactive obsolescence.
I’m here to help you
Currently, large language models can produce code roughly as well as an entry-level software engineer. If you give AI your favorite 20-minute interview coding question, it will probably solve it within a few seconds. AI is particularly good at translating code between languages and writing templated code.
To stay competitive as a software engineer, you should use AI for anything it can do well. That may include trying AI on every coding assignment you get to see which ones it nails and what ideas it suggests. AI can also refactor your code and make it more readable. Regardless, you should get comfortable using AI frequently. Otherwise, you’ll be left behind by your peers who do.
Since AI makes mistakes, you need to check its work carefully, just as you would for any entry-level engineer. Think of AI as your dedicated, speedy, entry-level, coding buddy. Yeah, maybe you can write code better than your buddy can, but you’d be crazy to have your buddy sit around doing nothing.
Eric Aside
For information on responsible use of AI-generated code, read Responsible AI Adoption—GitHub Resources. For more on code and design reviews, read Review this. For more on basic code quality, read Nailing the nominals.
It absolutely will not stop
As AI inevitably gets better, less of your job will be writing code, and more of it will be designing and reviewing code. In other words, much of your job will change from low-level work to high-level work.
If you love writing code and its instant gratification loop, imagining that part of your job going away might be disheartening. Get over it. For most software professionals, the days of manually writing new computer code are numbered. Instead, you’ll need to take joy in designing systems and services, specifying them clearly, and validating that the resulting AI-generated code meets your design goals.
Basically, software engineers will become clear, grounded, software architects. Clear because otherwise AI won’t follow your intentions. Grounded because otherwise AI won’t produce practical results. If you’re not an architect today, consider picking up those skills, particularly about being clear and grounded (many architects aren’t). Yes, it’s a different role, but it’s still fun, challenging, and creative.
Eric Aside
For more about being a great architect, read The other side of quality—Designers and architects.
Your son, Sarah
Eventually, artificial intelligence (AI) will graduate to artificial general intelligence (AGI). Some people doubt this, but I believe it’s realistic within the next decade. Regardless, there’s a broad consensus around AGI developing within our children’s lifetimes. People argue over exactly what AGI will be capable of, but it will at least match the capability of our brains. Where does that leave our kids?
If a computer can do any job a person can do, what jobs remain? The jobs people want people to do. Some examples include arts, trades, storytelling, socialization, child and personal care, mentorship, and companionship. Yes, robots and chatbots are already providing many of these services, and there are folks who like computer art or chatbot conversations. However, humans are social beings and thrive on human engagement, so interpersonal occupations will likely remain in demand. If that sounds too restrictive to you, try these roles sometime. They are rewarding for the giver and receiver on many levels.
Will future interpersonal occupations be rewarding financially? While the most skilled people (or most famous) will be compensated well, the rest will likely earn a living wage at best. Does this mean your kids will have a worse life than you? Not necessarily.
Of course, dystopias are possible and will exist in some places as they do today. However, governments are already trying solutions for when computers automate nearly all the high-paying jobs. These experiments, like universal basic income (UBI), admittedly have mixed results so far. That’s the point of experimenting. But automation brings down the price of goods and services, so when everything is automated, costs become negligible, and we all get to retire. This scenario assumes cheap energy, but AI is already driving investment in fusion, geothermal, and other abundant green energy. You might worry, but you might also be excited.
Eric Aside
For a brighter outlook on the world, including overpopulation, read Factfulness: Ten Reasons We’re Wrong About the World—and Why Things Are Better Than You Think. For more on working well with people, read You’re no bargain either, Evil assumptions, and Fuzzy logic—The liberal arts. For a sci-fi examination of how superintelligent AGI and humans might live side-by-side in abundance, read The Culture Series.
Your world… it’s pretty terrifying
The future can seem scary. So much is unknowable. But change is certain, and those who embrace it gain an advantage financially and psychologically.
Use AI as often as you can alongside your manual tasks, particularly for work it’s good at. Try it out even when you doubt the results: It could save you time or enhance your ideas. Carefully check AI’s output as you would for any entry-level engineer. Over time, as AI gets better, focus more on high-level design and review activities. Develop your skills as a clear, grounded, software architect. Finally, encourage your children to pursue careers that humans prefer other humans do, since it’s likely that in their lifetimes, computers will be capable of doing everything else.
The world can happen to you, or you can happen to the world. Embrace change and use it to your benefit. Take control of what you can and make the most of what life offers you and your family. That’s always been a winning strategy.
Special thanks to James Waletzky and Bob Zasio for reviewing the first draft of this month’s column.
Want personalized coaching on this topic or any other challenge? Schedule a free, confidential call. I provide one-on-one career coaching with an emphasis on underrepresented, midcareer software professionals. Find out more at Ally for Onlys in Tech.
> automation brings down the price of goods and services, so when everything is automated, costs become negligible, and we all get to retire.
Strong disagree.
We’ve seen lots and lots of technological upheaval in the last several hundred years (and far longer ago than that if you include the agricultural revolution and pre-industrial technologies). I’ve heard the promise of developer retirement within a decade multiple times in my lifetime alone.
Instead – what happens? Every single time there’s a massive leap forward, there’s a corresponding leap in expectations. People don’t just retire; instead, as costs come down, a larger number of people increase their standard of living significantly, and they adjust and continue working in a new world. The utopian-sounding “do whatever you want – you’re retired!” has been predicted for hundreds (thousands?) of years, but never actually happens.
There’s a possibility that AI may get to a senior level dev, but I strongly suspect not. I think we’ll start to see diminishing returns in the next year or two, and AI will settle in as just another tool among many.