GitHub CEO reveals the key to thriving as an engineer in the AI coding era

AI isn’t replacing you, it’s enhancing you

GitHub CEO Thomas Dohmke insists that artificial intelligence will transform how engineers work, not replace them. By automating routine tasks, AI allows developers to focus on more strategic, creative work.

This evolution means engineers can stay in their “flow state,” solving complex problems without constant context switching. Dohmke believes that developers who embrace AI as a productivity partner will thrive in this new coding paradigm.

Flow state is your new superpower

Flow state, the mental zone where you’re most productive, is central to Dohmke’s philosophy. AI tools like GitHub Copilot reduce distractions, letting developers dive deep into problem-solving without pausing for boilerplate code or syntax checks.

The goal is to remain focused and efficient, with AI managing the busywork. This creates more space for innovation and deep technical thinking, which are vital for thriving in modern software development.

Choose your agent like a pro

According to Dohmke, engineers must learn to select the AI agent that delivers the best return on investment, or code manually when it’s faster. Not all problems are best solved with AI.

Writing a few lines of code is sometimes quicker than crafting a complex prompt. The real skill lies in knowing when to use AI, which tool to pick, and when to trust your muscle memory.

Augment, don’t outsource your code

The biggest mistake developers can make? Treating AI like a black box. Dohmke says engineers should quickly take AI-generated code, tweak it locally, and keep building. You’re losing productivity if it takes longer to prompt AI than to write the fix yourself.

The best engineers use AI as a creative partner, not a replacement, refining and improving suggestions to suit their projects.

The rise of vibe coding, friend or fad?

“Vibe coding,” coined by OpenAI’s Andrej Karpathy, is making waves. It’s about leaning on AI tools to handle most of the coding so you can “forget the code even exists.” Dohmke is skeptical about relying solely on vibe coding.

He sees it as helpful for prototyping, but insists startups need deep technical understanding to scale and justify serious investor interest.

Startups can’t run on vibes alone

Speaking at VivaTech, Dohmke warned that AI isn’t enough to build a scalable startup. Non-technical founders might launch something cool using vibe coding, but without real developers, they’ll struggle to expand.

Investors are wary of teams with no coding depth, questioning whether such startups can handle complex systems or survive long-term growth stages. AI is a tool, not a full-fledged founding team.

AI won’t build investor confidence solo

Dohmke emphasizes that bootstrapping a startup with just AI might look innovative, but it won’t hold weight with serious investors. Differentiation vanishes when multiple teams use the same AI tools to create similar products.

What sets you apart is your technical execution, your product insight, and your capacity to scale. Without foundational skills, raising that next round of funding is tough.

Think like a developer, even with AI

The GitHub CEO believes prompt engineering is sound but not a replacement for fundamental programming skills. Writing prompts for something you could just code in seconds is counterproductive.

The goal isn’t to lean on AI constantly, but to know when and how to use it efficiently. Engineers should understand the logic behind what AI produces, ensuring quality and security in the final product.

The creative vs. production balance has shifted

Historically, software engineers spent much time on the “production” side of development, writing boilerplate code, debugging, and doing repetitive tasks. AI is changing that. Developers now have more time to design, experiment, and think strategically.

Dohmke compares coding today to composing music: the better your creative instincts, the more value you bring. AI just plays the instruments faster.

We’re climbing the abstraction ladder

Coding has continuously evolved through levels of abstraction from assembly to high-level languages. AI takes it further by letting developers describe their goals in natural language. Dohmke sees this as the next big leap.

You’re no longer coding line by line; you’re orchestrating ideas. But even at this height, human judgment is essential to shape what AI builds into something secure and scalable.

Copilot isn’t magic, it’s smart assistance

GitHub’s Copilot doesn’t aim to be autopilot; it’s more like a co-driver. It helps navigate common patterns, suggests fixes, and accelerates ideation.

Dohmke describes it as a partner that removes the friction of repetitive tasks. Developers are still in charge. The better your understanding of the codebase, the better you’ll direct Copilot to generate meaningful and secure contributions.

GitHub’s massive reach shapes the future

With over 150 million users, GitHub has a front-row seat regarding how AI is changing software development. Under Microsoft’s ownership since 2018, the platform continues to expand its AI tools.

Dohmke’s vision is clear: AI should amplify developer creativity, not replace it. GitHub’s ecosystem is now a proving ground for testing the boundaries of AI-human collaboration in real-world development.

Developers still define the direction

No matter how good AI gets, the human developer sets the course. Dohmke sees engineers as the visionaries shaping the product, making key architectural decisions, and ensuring usability.

AI doesn’t know your customer, roadmap, or technical debt. That’s where developers come in, bringing human context that AI simply can’t replicate. This is the essence of creative engineering.

Prompting is a skill, not a crutch

Prompting AI effectively is now a core developer skill, but it should be used strategically. Dohmke believes prompt literacy can increase productivity, but shouldn’t replace understanding core concepts.

Developers who rely too heavily on AI may overlook security, performance, or architectural issues. The most valuable engineers will blend prompt-crafting with technical depth to get the best from both worlds.

The AI arms race is just beginning

Dohmke believes we’re still early in the AI development journey. As models improve and integrations deepen, developers will need to adapt quickly.

Those who build alongside A, I treat it as a co-creator rather than a competitor, will shape the future of software. Staying ahead means staying hands-on learning, and embracing new tools with a thoughtful, strategic mindset.

Want to see how browsers are jumping in to help? Check out how Opera’s AI is stepping up as your next coding buddy.

The future belongs to hybrid engineers

Dohmke’s closing message is clear: the future belongs to engineers who blend technical skill with creative vision and AI literacy. You’re unstoppable if you can write great code, direct an AI agent, and think like a product strategist.

AI is changing how we code, but not why. The Thrive developers will embrace the shift while staying grounded in what makes their work uniquely human.

Curious how other AI tools are leveling up for devs? See how Google just made Gemini smarter with GitHub code.

What do you think about the tips the CEO of AI gave to be on top in the AI industry? Please share your thoughts and drop a comment.

Read More From This Brand:

• Trump Team’s Secretive AI Agenda Allegedly Exposed in GitHub Leak
• GitHub Copilot Adds AI Tools, Limits Premium Access
• Microsoft Layoffs Hit Programmers Hard As AI Writes More Code

Don’t forget to follow us for more exclusive content on MSN.

If you liked this story, you’ll LOVE our FREE emails. Join today and be the first to get stories like this one.

This slideshow was made with AI assistance and human editing.