This question has been at the center of heated debates on LinkedIn over the past few months, and we wanted to offer a more thorough discussion with experts in the field.
The panel featured three seasoned tech leaders:
So how is AI currently transforming the role of developers? What is the impact on hiring practices and internal training? And perhaps the most pressing question of all for junior developers: What should the next generation of programmers focus on?
Let’s discover what our panel had to say!
With this first question, we wanted to look under the hood—past the clickbaity headlines—and understand how AI-powered tools are actually changing the way software engineers work.
Chris described how her team at Google is encouraged to use AI tools every day: “Whether it’s for coding, ideating, or even for kids’ homework. Just use it and get familiar with what it can and can’t do.”
For code specifically, she gave a great example: “Obviously we use Gemini a lot. Just today, I was checking some regular expressions and writing tests for specific use cases. Even if it can’t do it perfectly, it can do it really, really fast compared to what I could do before.”
We also learned from Masa about how AI completely transformed how Progate engineers their products: “It used to be that 90% of a developer’s job was writing code. Now, it’s more about orchestrating, reviewing, and translating ideas into architecture. AI can review your code, list out things that you’ve left out. It does the typing part for you.”
So has the skill set required for developers changed? “Obviously, engineers now need to develop their prompting techniques,” Boris explained, “The execution part has been commoditized. What matters now is how clearly you can define what needs to be built – specify the feature. You basically need to get into a discussion with your AI assistant.”
Before concluding on that first question: “We’re seeing a new breed of junior developers who have a very strong product mindset. Product managers are also turning into builders, specifically for launching MVPs or quickly testing features.”
Interestingly, a second question (which, as a moderator, I did not plan for!) naturally came up: What is the impact on junior developers specifically?
Masa put it bluntly: “In the first two years, it used to be about doing little tasks. Now, you have to prove to senior engineers that you can deliver more than Devin. Simple markup tasks alone no longer justify a junior developer’s position.”
And Chris herself was no less direct: “Obviously you need to learn the AI stuff, and you will fall behind if you don’t learn how to use those tools effectively. That is true for anyone, senior and junior.”
She also shared a personal experience illustrating that: “It’s a lot easier to skip as a developer. The whole ‘vibe coding’ thing is great for prototypes, but dangerous for large-scale applications. I was building a small Android app with AI, and it kind of worked… but I could tell it was going to be very bad. I had no idea whether I had 20 times more functions than I should have—or what any of them were doing.”
Her conclusion? “As a junior developer – or someone learning a new language – you need to be more intentional about that process: go through tutorials, read books. Take the time you need to actually learn.”
“Google hasn’t necessarily changed their interview practices. You can’t use AI, and it’s considered ‘cheating,’” explained Chris.
But she noted how other companies are adapting: “Some companies have moved to full on-site interviews. Others are testing your ability to build an entire application. They’ll ask you to walk them through your approach. With AI tools, you can do bigger things in interviews now – which is exactly what I was lacking with Android, for example.” (laughs)
Boris added: “I think there are two aspects to it. First, how you leverage AI-powered tools as a developer. Second, what’s your experience playing with AI-related apps – LLM API calls, prompt versioning, the issues you run into when developing an agent.”
We also wanted to hear from Masa how Progate encouraged their teams to adjust to AI: “Back in December last year, we made Devin available to all our engineers. But after two weeks, we realized there were ZERO pull requests using it. So we made it mandatory.”
And it worked! “That push was very necessary I think. We’re now seeing more pull requests using Devin.”
Each speaker offered advice for developers trying to future-proof their careers, and we compiled these in a to-do list for you!
All panelists agreed: developers who know how to prompt well, debug generated code, and use AI to augment their thinking will stand out. Whether it’s Cursor, Claude, Bolt, v0, Gemini… You have to try things out.
Knowing when code is “smelly” or unscalable only comes with experience: “You can ask AI to build something,” said Masa, “but only a developer can tell if it’ll break with a million users.”
“New tools come out pretty much every month, and you’ll have to learn how to use them,” Chris said. “Your workflow today might be obsolete in six months. Stay open-minded and keep learning.”
Lastly, Boris insisted on the broader product skills: “Build your own experience on what it takes to develop an AI solution. It doesn’t have to be a complex solution or a system of agents – just an elegant use of AI to solve an actual problem.”
After a round of audience questions geared towards junior developers and how to navigate the tougher job market, the panel came to a shared conclusion:
No, AI won’t replace developers. But the job will look very different.
Developers will need to focus more on architecture, product thinking, evaluation, and cross-disciplinary communication. New hybrid roles, such as AI evaluators, prompt engineers or agent designers will emerge.
Chris summed it up: “Everything will become an AI product. That means everything we know about building software needs to evolve.”
Masa added: “Self-teaching will become easier. But the gap between those who truly want to learn and those who rely too heavily on shortcuts will grow.”
And Boris provided a final conclusion: “There’s no free meal. AI is powerful, but without understanding the fundamentals, you’re just creating chaos faster.”
_______________________
Special thanks to our panelists Chris, Masa and Boris, as well as Google for Startups Tokyo for welcoming us!
When you think about ownership, it’s more than just having a thing. It’s also taking
Learning Python in today’s tech-driven world is like equipping yourself with a powerful weapon. This
The global technology market is evolving and growing. Emerging technologies in the way of artificial
Becoming a developer is not the only career path after a Coding Bootcamp. For Carole,
Attracted by analytics and the top, in-demand careers, Robyn had one objective in mind: to
So you’re thinking about joining our coding bootcamp but are unsure because of COVID19? In
