This post was originally published on this site.
In fact, we recently posted some internship positions, and within one day, we got around 700 applicants. That never happened before. I think that’s due to the current narrative. People are still trying to figure this all out.
But I think it will become clear that learning to code is a fundamental skill—like learning math. Even though we use calculators and spreadsheets now, we didn’t stop teaching math. It’s the same with coding.
Yes, AI tools can generate effective code, but you still need engineers who can use that code properly—who think logically, solve problems, and understand how systems work. And if you don’t hire junior engineers, how will they gain experience? How will your organization grow? It’s all connected.
I think there’s also a mindset shift happening. Even junior engineers need to start thinking like team leads or dev managers, because they’re essentially managing a team of AI agents. They’re planning tasks, breaking them down into smaller pieces for the AI to handle, and then reviewing the results.
That’s definitely a challenge, especially for junior engineers who are getting code from tools like Tabnine—how can they trust what’s been generated if they don’t fully understand it? It’s a mindset shift that applies across the board, but senior engineers are more prepared for it.
The new expectation is that even entry-level engineers need to start thinking this way. Are most engineers you talk to adopting this mindset? Are they embracing the technology, or do they still see generative AI as a threat?
We had a similar fear on the media side, especially with writing—AI generating articles, blog posts, and so on. But I think most of us now have accepted it. We use it as a starting point and refine from there.
A lot of the AI-generated content—especially for things like social media posts—is actually pretty good and helps reduce repetitive tasks. Is it the same on the coding side? Or is there still fear or resistance? Eran, we’ll start with you. No, I don’t think there’s an adversarial relationship—quite the opposite.
People are rooting for these tools to succeed and pushing their boundaries. Just like with writing, the people who get the most value are the ones who’ve figured out where the tools work well and where they don’t.
Good engineers know when to use AI and when to write code themselves. They’re developing a kind of intuition—knowing when something can be automated and when it can’t. The good engineers are embracing this and using it effectively.