Is AI coming for my job?
For the time being I make my living as a software developer, the last ten years my career have been great: it's provided me good pay, job stability and afforded me a pretty cushy lifestyle. Lately there's been an obvious industry shift (I'm talking about AI) that makes me nervous these things I took for granted may not last forever.
In the last couple years, AI assisted coding has completely changed I write software. In March 2025 the Anthropic CEO Dario Amodei predicted that AI would write 90% of software code within 3 to 6 months and essentially all code within 12. At the time I scoffed at the claim, but now in 2026 can't remember the last time I fired up my editor of choice (vim btw) to write out code by hand.
I'm equally parts concerned and amazed that I can comfortably hand off a ticket or issue to an LLM and have it come up with an plausible solution. A year ago, the results were much worse, with every new model release the results continue to improve. A year from now, who knows how much better the frontier models will be?
As someone who's always prided themselves in being a programmer this makes me pretty nervous. I've always taken pride in my work, love programming and (still) care about the art and craft of software development.
Understandably I'm concerned about a future where the value of individual software developers is completely undercut by AI. If you're a company, why pay an expensive developer to build/maintain software when you can just use AI. The trade-off might not make sense today, but for how much longer (if at all) does that remain true. Remember developers aren't the ones making this trade-off, it's management.
I'm still a way off becoming a complete meat proxy but it's quite scary how often I'm just approving and merging LLM authored code with minimal modifications. If that's all you're doing now how can you keep justifying the same salary for your $EMPLOYER?.
In January 2026 Anthropic CEO Dario Amodei predicted that AI models could handle "most, maybe all" end-to-end software engineering within 6 to 12 months. This poses a clear morale dilemma. In the short term these tools feel like they make my life easier, and make me more productive. But why would I continue to pay Anthropic for a monthly subscription when they so brazenly claim that AI tools will likely wipe out my job in the future?
I'm still dubious about these claims. I think this (and most of Anthropics public coms) are just marketing stunts to pump the stock price before IPO. We're currently in October 2026 and software engineer certainly doesn't feel 'solved', if anything the overall quality of software feels like it's continuing to degrade.
Software being 'solved' aside, perhaps I'm just caught up in the hype, but when I see the pace that things are changing, it feels like either of the following two scenarios are inevitable:
- Programmers get laid off
- The role of programmers changes
Layoffs aren't anything new in the world of software. During COVID companies clearly over-hired and there's been some brutal layoffs in the subsequent years as a result (leadership will use AI as a convenient excuse). I'm lucky I've never been personally affected by layoffs, but the current transformation the industry is undergoing feels more existential for the entire profession in general, not just individuals.
It's easy to feel like I've painted myself into a corner the way I've carved a career out for myself. I've always been a 'developer' and have historically resisted pushes towards management, product or leadership style roles. In hindsight I think that may have been a mistake.
Despite my role I've never considered myself a 'cracked' programmer, I've always struggled with leetcode style interview questions, math and low level programming. If you want a CRUD API that communicates with external services I've always been your go to guy, but this is the exact type of work AI really excels at.
What I'm Doing About This
Instead of feeling helpless I want to try and shape my own destiny. This is tricky because historic career advice in the field of software development was pretty straightforward, so there's no precedent for what's happening.
I'm guessing there's a couple things I can do here. The first is I'm going to try my best (where possible) to get more involved in customer/client facing work. Being a technical point of contact with external clients (to me) feels like the kind of thing AI is never going to be able to replace. Maybe there's some #ShittyFuture where an AI reaches out to another business to discuss requirements but I don't see that happening anytime soon.
The second is I want to double down and really embrace these AI tools to make myself as efficient as possible. I'm not talking about spawning fleets of autonomous agents to ship 20,000 commits of slop every 24 hours, but I'm talking about utilising AI in a pragmatic fashion to output higher quality software (something I think not enough people care about).