At the second annual AI Dev summit in New York City, hosted by DeepLearning.ai and its founder Andrew Ng, the future of AI in software development was the main topic of discussion. The event gathered industry leaders, developers, and AI experts to talk about how artificial intelligence is transforming coding, employment, and the broader tech landscape. In several panels and a dedicated interview with ZDNET, Ng shared his insights on how professionals can adapt to the rapidly evolving AI landscape, the importance of AI literacy, and the role of regulation in AI’s future.
### The Changing Landscape of Coding Jobs
The integration of AI into software development has already begun to reshape the industry. AI-powered coding assistants are now common tools for developers, automating many routine tasks that used to be handled by entry-level programmers. As a result, there are fewer junior coding jobs available for recent graduates, raising concerns about the future of the profession. However, Ng and other experts at the summit argued that despite these changes, engineers will not become obsolete. Instead, the profession is evolving, with new skill sets and roles emerging in response.
Ng believes that the barrier to entry for coding has been dramatically lowered by AI. He advocates for a broader approach to coding education, suggesting that everyone—not just traditional software engineers—should learn the basics of coding with AI assistance. He compares this to knowing a bit of math: a fundamental skill that can be applied across many careers, regardless of whether someone becomes a full-time developer. In Ng's view, the essential skill is being able to “tell a computer exactly what you want it to do,” emphasizing understanding over memorizing syntax or technical details.
### The Rise of “Vibecoding” and Generalists
At the summit, Ng introduced the concept of “vibecoding” (which he prefers to call “AI coding”)—the practice of using AI tools to generate code, even for those who aren't professional developers. He welcomes non-traditional coders into the community, recognizing that AI is making software creation accessible to more people. However, he also acknowledged the mental effort involved, saying that working with AI tools can be “mentally exhausting.”
A recurring theme at the summit was the growing need for professionals to become generalists. With AI making the technical side of software development faster and more efficient, the bottleneck in product creation has shifted from coding to product management and conceptualization. Ng advised engineers to learn product management skills, arguing that those who can handle both technical and strategic aspects can function as highly effective “teams of one.”
Panelists highlighted that domain experts in non-software fields can now use AI coding tools to rapidly prototype and iterate on ideas, breaking down barriers between disciplines. The challenge for future developers, Ng noted, will be defining the right problems and solutions—coming up with new concepts rather than just writing code. Fabian Hedin, CTO of Lovable, echoed this idea, suggesting that developers will need to rely on intuition and a deep understanding of