Is software still a career?

A few months back I wrote a comment on LinkedIn to a young person who was giving up on learning to be a software professional - I suggested that it was to soon to give up on software - and that AI only made software more relevant, in that the reach is going up with everything we can do now and the speeds we can work at. There's going to be software for everything* - what a great time to jump on the train.

It's easy to look at what's happening in AI and see the end of humans in it - but as someone working in the industry there's more nuance: Yes - organisations are changing and there'll be less boiler room coders in a typical software product organisation. Instead there'll be a massive growth opportunity for product engineers. What's a product engineer - it's a person with the ability to reason about why the software is being made not just how to make it. AI is a tremendous opportunity if the why of software is as interesting to you as the _how.

My comment somehow struck a nerve - because a few months later I was asked again whether I thought it was a good idea to take an education in software engineering. This time it was someone a little further along in their career thinking about transitioning into software.

And obviously this is the $10000 dollar question we're all grappling with and I don't know the answer - but if you're someone who is curious about the world it's never been easier to explore it through software than now. That's not a problem, that's an opportunity. Perhaps you'll not be hired into an organisation with thousands of engineers - but that's probably a good thing too.

What you can make read from what I've said above - I actually think coming from another field is an excellent position to be in when entering into the software field - domain expertise is not going away - if you educate yourself at the intersection of engineering and something else, you're a natural fit for a life as a product engineer.

Obviously the engineering schools need to take AI in - and fast - if they're teaching algorithms maybe they should rethink the curriculum a little bit.

*As an aside: Maybe a third of the demos at last week's Demodag in Copenhagen were made for audiences of 1-10 people; we can serve these needs at low cost now