💬 Your voice matters and now it comes with a little perk! Introducing Question of the Month

Hi all!

Hope you had a great weekend 😊

From this month onwards, we’re starting a new tradition in the Community: Question of the Month. Each month, we’ll ask one thoughtful question for everyone to share their ideas, experiences, and perspectives.

It’s all about sparking conversations and learning from each other. And as a little thank-you for joining in, we’ll randomly select one participant each month to receive a £25 “Treat Yourself” voucher to spend on what you decide 🎁

✨ This month’s question

This month we have a question from one of our brilliant automation consultants, @Maya Vashi

Is learning to code still a career differentiator - or will AI make it optional?
AI copilots can generate working code from plain language, shifting what “coding” means. We’re curious: over the next 3–5 years, will coding remain a core capability for professionals, or will higher-order skills (problem framing, prompt design, governance) matter more?

What do you think? We'd love to hear your take on it, share your answer in the comments below to enter this month’s draw - Terms & Conditions apply

Jump in the conversation!

It looks like you're new here. Sign in or register to comment.

Find more posts tagged with

Question of the Month

Comments

AJT

Full disclosure, I’ve never been a programmer so I’m ignorant to development best practices/methodologies/productivity tools. I’d expect AI to support rapid code development but only produce usable output when used by a sufficiently skilled developer, who already knows the underlying principles of a chosen language and a good grasp of the underlying codebase (if extending existing software). It also feels like the more complex the request, the more likely it is that it will require manual intervention to refactor, which might be very difficult or impossible if the prompter doesn’t understand the fundamentals.

Using myself as an example, I could probably get a simple bit of code working after a few prompts but I’d have little confidence in being able to effectively debug or fully understand how the code might have unintended effects elsewhere. I’d consider it reckless to build any kind of codebase without really understand what I’m doing. The higher-order skills are important, but not nearly as important as understanding how to code, a language’s best practices, software architecture, debugging/troubleshooting/maintaining, and ultimately knowing what the software should be doing.

I imagine that the AI wave will continue and AI tools for development will become more prevalent. I can also imagine an incredible amount of technical debt being accrued if no experience / low experience developers start creating AI code in volume without any oversight or review. As with any cycle, I can see it coming full circle and experienced developers being in high demand to fix underlying problems caused by AI coders.

Either that or AI tools will progress so rapidly that it renders all of the above moot...

Maya Vashi

Thanks for sharing your thoughts @AJT - I agree with a lot of it.

I do come from a programming background, and even with that experience, I’ve found that AI tools are best used as accelerators, not replacements. I think they're great for scaffolding ideas, generating boilerplate, or even suggesting optimisations - but again, only when you already understand the language, the architecture, and the implications of what’s being generated.

You nailed it with the point about complexity. The more intricate the task, the more likely it is that AI will produce something that looks plausible but needs serious refactoring. And without a solid grasp of the fundamentals, that refactoring can be risky or even impossible!