• OOP Programming made designing code easier.
• Java made learning it easier.
• IDEs made writing it easier.
• Cloud computing made running it easier.
• Containers made building it easier.

AI is different. Even though prototyping has never been easier, building maintainable, robust systems over the long term is still far from simple. What AI does go far beyond all of that – it’s redefining what it means to be a software developer. And as any major change, there is a good part, a bad part, and sadly, an ugly one.

The Good

• AI makes bad developers worse and good developers better. The gap between them is becoming more visible—and may widen into an irreconcilable chasm. Not necessarily a bad thing but a complex organizational issue.
• AI helps me focus on the skills I enjoy most such as writing clearly, understanding design principles and patterns, and iterating quickly on product features.
• AI helps me focus on what to build rather than how to build it. AI is a thinking partner and my favorite rubber duck during debugging.
• AI makes side projects finally come true, encouraging experimentation and creativity.
• AI is a great learning tool. AI is a problem-solving tool (give me the answer) AND a learning tool (explain me the answer). I’ve never asked so many questions that I couldn’t articulate so clearly before. I even sometimes ask AI to test my comprehension!
• AI challenges the recent specialization of developers by making full-stack development more accessible again, closer to how it used to be.
• AI can improve code quality despite concerns about technical debt. With the right developers, AI can finally address the long list of TODOs and bugs that have been sitting in backlogs for years.

The Bad

• AI has a significant environmental cost—arguably higher than most previous technological advances.
• AI raises the bar for junior developers.They now compete with powerful AI agents that have literally read all GitHub public repositories. Not all students will reach the minimum level to meet the new expectations. What about professional developers that aren’t passionate (not a judgment)?
• AI often lacks originality. Imagine browsing bookshelves where every book sounds similar. That might be acceptable for code—but what about documentation, blogs, or podcasts?
• AI limits innovation (in some aspects). AI excels on proven technologies but may miss newer ones (the knowledge cutoff). If developers increasingly rely on AI to design systems, the incentive to create new frameworks or libraries may weaken. Similarly, why write blog posts that train AI models if they won’t properly credit you? Over time, motivation may fade away.

The Ugly

• AI is disrespectful. Models were trained on a massive volume of human creativity: OSS projects, websites like Wikipedia and StackOverflow, books, blogs, etc. Sam Altman recently tweeted to express his gratitude to developers¹, except OSS developers have never written code publicly with that goal in mind. Why not also thank authors whose books were pirated from shadow libraries. Or Wikimedia benefactors that pay for the infrastructure used by AI models to absorb human knowledge.
• AI jeopardizes the labor market, starting with developers. Oracle recently fired 30,000 employees overnight² and those massive layoffs are just the first iteration of what work will become tomorrow. And governments aren’t prepared for that. How to prepare for the unexpected when it’s already so hard to prepare for the expected such as climate change.

There is much more to say—and even more that I don’t yet understand. This is a complex topic, and these are just my reflections, not definite conclusions. As Derek Thompson expressed so well in a recent blog post³ about where AI is going:

I am lucky to have participated in conversations about the future of AI with executives and builders at frontier labs, economists at AI conferences, AI investors, and other bigwigs at off-the-record dinners where important truths can theoretically be bandied about without risk. And if I had to pick three words to summarize this collective expert view of the future, I could not in a million years, or with a trillion tokens, find three words more suitable than these: Nobody knows anything.

In short, we don’t know where we’re going—but we’re going anyway.

---

Footnotes

1. https://x.com/sama/status/2033935276079510011 ↩

2. https://www.ibtimes.co.uk/oracle-layoffs-30000-employees-6am-email-1789644 ↩

3. https://www.derekthompson.org/p/nobody-knows-anything ↩