
Microsoft AI CEO Mustafa Suleyman said that building software has become so easy that anyone can now create a functioning web app “in seconds” with Vibe Coding.
Mustafa Suleyman on Vibe Coding’s Potential
Vibe coding refers to the practice of using AI tools to generate code and applications through conversational prompts rather than traditional programming.
It has been the biggest discussion point among software engineers and companies in the last couple of years. Now, one of the biggest tech founders has shared his thoughts on it.
In a recent episode of the Exponential View podcast with Azeem Azhar, Microsoft AI CEO Mustafa Suleyman (who is also the co-founder of DeepMind) said:
“It’s about vibe coding with it. It is so accessible now. You can watch a 3-minute video, get spun up, launch one of these things, and suddenly you are watching an AI code up some idea that you had for organizing your family calendar or, you know, planning out your workouts for the weekend or whatever it is.
And you can create an app, a web app in seconds, and learn, you know, by watching and by doing and prompting an AI engineer to go and build something that you may have thought was never possible.”
What he is really saying here is that building software no longer feels like coding in the traditional sense.
The “watch a 3-minute video” part is important. It highlights how low the entry barrier has become.
Earlier, learning to build an app could take months. Now, you can watch a short tutorial, open an AI coding tool, and immediately start creating something real.
This shift has major consequences:
- Creation is faster and cheaper: Traditional software development often takes weeks or months. With vibe coding, prototypes can appear in days or hours, and simple apps in minutes.
- Competition increases: When anyone can create software easily, markets become more crowded.
Then, Mustafa shared how he used Vibe Coding recently:
“I have definitely got one that tracks all the music things that I love, like when various DJs are playing, when concerts and festivals, and like then tries to plan out to map that to my travel schedule.”
They also discussed the societal risks of conscious AI and the potential dangers of people projecting consciousness onto AI systems.
So, he vibe-coded a system that tracks his favorite DJs, upcoming concerts, and travel schedule. This automates what used to be manual work into a self-updating spreadsheet.
The Problem with “Vibe Coding”
While Suleyman champions vibe coding’s accessibility, others in the industry are looking beyond the hype to what comes next. Bret Taylor, OpenAI’s board chair, acknowledged in a “Big Technology Podcast” interview that building software quickly through vibe coding will soon feel routine.
Taylor believes that AI agents will replace traditional software interfaces. Instead of navigating dashboards or filling out browser forms, users will increasingly interact with autonomous AI agents that execute tasks on their behalf.
That seems a little too far-fetched.
Mustafa’s take on vibe coding sounds exciting, but it also glosses over some very real problems.
Yes, AI makes it easier to spin up an app quickly, but “working” doesn’t mean “working well.” Most vibe-coded apps are fine for demos or personal use, but they often fall apart when real users, real data, or real scale show up.
Performance issues, security holes, and messy logic are easy to hide when an AI is doing the heavy lifting and the human doesn’t fully understand what’s under the hood.
There’s also a risk of overselling accessibility. Watching an AI write code feels empowering, but it can create a false sense of mastery. If something breaks, many vibe coders won’t know why or how to fix it without asking the AI again.
Bottom Line
Marc Andreessen famously declared that “software is eating the world” in 2011. Fifteen years later, we’re witnessing the next chapter: AI is eating software itself.
For innovators, it presents an extraordinary opportunity. The same tools threatening established companies also enable new entrants to compete on ideas rather than technical resources.
The implications of this shift became painfully clearwhen Anthropic announced legal-focused capabilities for its Cowork assistant. The AI tool can now review legal documents and handle compliance work. It caused a panic in the world and was dubbed as the “SaaSpocalypse.”
Other than Vibe Coding, Suleyman advocates for ensuring that AI development remains aligned with human interests and is controlled by humans, rather than seeing AI as a natural evolution beyond humanity.
Table of Contents
Related articles

Karpathy Says LLM Coding Split Engineers But Misses the Real Cost
Andrej Karpathy predicts LLM coding will split engineers into builders and coders. But skill atrophy and quality issues are also challenges.

Sam Altman Exposes AI Washing in Corporate Layoffs
OpenAI CEO Sam Altman calls out companies for “AI washing,” accusing them of blaming AI for layoffs that may not be AI-driven.

Snap Cuts 16% Workforce aiming $500M Savings with AI
Snap is laying off 16% of its workforce as it leans on AI tools to reduce costs and move closer to consistent profitability.

Is Labor Day Still Relevant in 2025 When the Labor is AI?
From fighting for eight-hour workdays in 1886 to battling job automation in 2025, workers face a new challenge. As AI replaces millions of jobs, the question isn't whether machines can do our work better. It's whether we can preserve human dignity in a world where algorithms never take breaks.


.avif)

