Something remarkable is happening across the world right now. People who have never written a line of code—accountants, teachers, small business owners, freelancers—are able to build an app. Not learning to build one. Not taking a coding bootcamp. Just... building one. With AI.
*How AI democratised product building, and why I believe we're about to witness the most chaotic innovation boom in tech history
*
Something remarkable is happening across the world right now. People who have never written a line of code—accountants, teachers, small business owners, freelancers—are able to build an app. Not learning to build one. Not taking a coding bootcamp. Just... building one. With AI.
"I need to build an application for tracking my daily tasks," they say, dead serious, while holding ChatGPT conversations like they're technical co-founders.
And here's the kicker: Some of them are actually succeeding.
Welcome to 2025, where the barrier to building software products hasn't just been lowered—it's been absolutely obliterated with the force of a $100 billion funding supernova.
The new .com craze?
Let's start with some absolutely bonkers statistics that sound like they plucked out of the air on a whim:AI funding exploded to over $100 billion in 2024—an 80% increase from 2023's already eye-watering $55.6 billion. Nearly one-third of all global venture capital is now flowing into AI companies. That's not just a trend; that's a tidal wave that I firmly believe is going to reshape the entire tech landscape.
But here's where it gets interesting and just that tiny bit loopy... code generation startups are commanding huge valuations. Cursor, the application I spoke about a lot in previous articles- with just 60 employees, went from zero to $100 million in recurring revenue in less than two years and just raised $900 million at a $10 billion valuation. Meanwhile, Windsurf launched its coding product in November 2024 and is already pulling in $50 million in annual revenue - it's batshit, scary and shows a lot about the predictions for the future.
These aren't anomalies—they're the new normal in a world where anyone can become a software builder.
The great democratisation
When I began web design it was simply HTML (and soon afterwards CSS and JS came to join the party). But more recently, tech stacks have become huge, often convoluted monstrosities - meaning that software development has been protected by a moat of complexity. Developers needed to understand multiple programming languages, frameworks, databases, deployment pipelines, and approximately 47 kajillion different ways that their code could break in production.Those days are deader than Internet Explorer 6.
Today's AI-powered development tools have transformed coding from an esoteric art form practiced by caffeinated wizards into something approaching normal human conversation:
Before AI: "I need to learn JavaScript, React, Node.js, Express, MongoDB, Docker, AWS, and sacrifice three weekends to the debugging gods."
After AI: "Hey Claude, make me an app that helps people find parking spots, but make it look professional and not like a geocities fever dream."
The AI doesn't just write code—with decent rules, MCP servers and small increments it writes great code. It handles the boring stuff (error handling, database connections, security considerations) while you focus on the actually interesting parts: solving real problems for real humans.
The WordPress moment: my prediction for the near future.
Remember when WordPress democratised website creation? Suddenly, every small business owner, blogger, and ambitious teenager could have a professional-looking website without hiring a developer or learning HTML. WordPress didn't just lower the barrier to entry—it completely eliminated it for 90% of use cases - now powering more sites on the internet than any other platform.**AI development tools are about to do the same thing, but for entire applications.
**
Just like WordPress spawned millions of cookie-cutter sites (how many "About Us" pages have you seen with the exact same layout?), we're about to see an explosion of AI-generated applications that follow similar patterns. The difference? These won't just be static websites—they'll be fully functional apps with databases, user authentication, payment processing, and all the bells and whistles.
The WordPress Parallel:
- WordPress: "I need a website" → Pick a theme → Add content → Launch
- AI Development: "I need an app" → Describe the functionality → AI builds it → Launch
And just like WordPress created an entire ecosystem of themes, plugins, and services, AI development will spawn its own ecosystem of templates, AI assistants specialised for different industries, and deployment services.
Within 18 months, I predict every freelancer, small business, and side-hustler will have an AI-built web presence that goes far beyond static websites. These will be interactive, functional applications that would have required a full development team just three years ago.
The explosion of Apps & Slop
We're about to witness what I'm calling the "Big bang of apps"—a rapid diversification of software solutions coming from the most unexpected places. When the technical barriers disappear, you get innovation from angles the traditional tech industry never considered.The Amazing Part:
Your child's after-school activity organiser who always complained about task management software? They're building something better. The accountant who got fed up with expense reporting tools? They're three weeks into developing an alternative. The barista who has opinions about inventory management? They're halfway through an MVP that's already getting interest from other coffee shops.
These aren't just hobby projects. Real businesses are emerging from people who previously thought "I could never build that myself."
The Terrifying Part:
We're also about to be absolutely buried under an avalanche of AI-generated slop. For every brilliant solution from an unexpected innovator, we'll get seventeen apps that solve problems nobody knew they didn't have:
"Uber for plant watering"
"Tinder for choosing what to eat for lunch"
"LinkedIn for your pets"
"Blockchain-powered sock matching" (because of course someone will add crypto)
The app stores are about to become digital flea markets where finding quality apps becomes an archaeological expedition.
###The new tech stack: Taste, Timing, and Tenacity
Here's the plot twist that traditional software companies are just starting to realise: when everyone has access to the same AI superpowers, technical skill becomes irelevant. The new differentiators aren't about who can code faster or debug more efficiently.
Taste and UX becomes the ultimate technical skill
The winners in this new landscape won't be the most sophisticated prompt engineers. They'll be the people who can distinguish between "this solves a real problem" and "this is just another app that reminds people to drink water."
Good taste in 2025 means understanding:
Which problems are actually worth solving
How humans really behave (not how we think they should behave)
The difference between "technically possible" and "genuinely useful"
When to build something simple vs. when to add complexity
Market timing still matters (maybe more than ever)
With development cycles compressed from months to weeks, market timing becomes even more critical. The first AI-built solution to market might not be the best, but it could capture enough attention to make the "better" version irrelevant.
Execution and iteration speed
When your main competitive advantage isn't technical complexity, everything comes down to how quickly you can listen to users, adapt, and improve. The ability to rapidly iterate based on real feedback becomes the ultimate moat.
The investment frenzy: When VCs start throwing money like confetti
The venture capital world has completely lost its collective mind, and honestly, it's entertaining to watch from the sidelines.Meta is reportedly in talks for a $10+ billion investment in Scale AI. SoftBank is committing $15-25 billion to AI infrastructure projects. Google launched an "AI Futures Fund" that's basically "take our money and access to our best models, please just build something cool."
Even traditionally conservative sectors are getting AI-drunk. Healthcare AI attracted $5.6 billion in investment in 2024. Financial technology startups with AI components are commanding premium valuations. Y Combinator's latest batch looks like an AI startup catalog.
The investment thesis seems to be: "If it has AI in the name and solves any problem whatsoever, here's a cheque."
This creates a fascinating dynamic where:
There's abundant capital for AI-powered solutions
The barriers to building those solutions have never been lower
The competition to find genuinely differentiated companies has never been higher
It's like a gold rush, except instead of pickaxes and pans, everyone's armed with large language models and unlimited optimism.
The quality control problem: separating signal from noise
Here's the uncomfortable truth nobody wants to talk about: democratising creation doesn't democratise quality.When professional gatekeepers (developers, designers, product managers) are no longer required to build software, we lose their quality filters too. The result? A lot more products, but with a much wider quality distribution.
The Good Signal:
- Innovative solutions from domain experts who understand problems intimately
- Rapid experimentation and iteration on novel ideas
- Products built by people who are actually experiencing the problems they're solving
The Overwhelming Noise:
- Solutions in search of problems
- Apps that work technically but ignore basic user experience principles
- Products built by people who've never talked to their target users
- The inevitable tsunami of AI-generated content farms disguised as "apps"
What this means for traditional tech companies
If you're a traditional software company watching this unfold, you're probably experiencing some combination of excitement and existential dread. And you should be.The Threat: Your proprietary technical advantages are evaporating. If a domain expert can build 80% of your functionality in three weeks using AI tools, your moat might be more like a puddle. The problem now is defensibility!
The Opportunity: You can accelerate your own development dramatically. That feature roadmap you planned for 2026? You might be able to ship it by summer.
The Strategic Question: Are you competing on technical execution or on understanding customer problems? Because if it's the former, you might want to pivot quickly.
The future: my own crazy predictions
Based on current trends and a healthy dose of speculation, here's what I think happens next:Short Term (Next 12-18 Months):
- App stores become increasingly cluttered with AI-generated products
- AI-generated sites become the new WordPress—every small business, freelancer, and side-hustler will have an AI-built web presence within 18 months
- The first wave of AI-built unicorns emerge (probably in unexpected verticals)
- Traditional software companies either embrace AI tools or get disrupted by companies that do
- LLM's become able to build flawless code and hold huge amounts within context.
Medium Term (2-5 Years):
- Market consolidation as the best AI-built products acquire users and funding
- New quality signals emerge (human curation, community validation, outcome-based metrics)
- AI development becomes sophisticated enough that non-technical founders can build genuinely complex products
- The definition of "technical founder" evolves completely
Long Term (5+ Years):
- The distinction between "traditional" and "AI-built" software disappears
- Product development becomes primarily about problem identification and user experience design
- We probably look back at this period as the moment software development became truly accessible to everyone
The bottom line: What should you actually do?
If you're reading this and thinking "I should probably build something," you're not wrong. But before you open ChatGPT and start prompting your way to unicorn status, consider a few things:Start with the Problem, Not the Solution The democratisation of building means the bar for execution is lower, but the bar for having something worth building is higher. What problem do you understand better than anyone else?
Embrace the AI Tools, BUT don't let them drive AI is incredibly powerful for implementation, but it's terrible at strategy and taste. Use it to build what you envision, not to decide what to envision. Trust me when I say that if you let them just plow on without you taking control - they will hallucinate and they will create terrible bugs.
Move fast, but think about quality The ability to iterate quickly is your superpower, but shipping garbage quickly is still just shipping garbage. Speed without quality is just noise.
Focus on humans, not technology In a world where everyone can build technology, the people who understand humans will win. Spend more time talking to potential users than prompting AI models.
My final thought: welcome to the wild west of software
We're living through the most dramatic democratisation of technology creation in history. For the first time ever, having a good idea and basic determination might be more valuable than technical expertise.This is simultaneously the most exciting and most chaotic time to be involved in technology. The barriers are down, the funding is flowing, and the field is wide open for anyone with a genuine solution to a real problem.
Just remember: in a world where everyone can build software, the winners won't be the best builders—they'll be the best problem-solvers who happen to use building as their tool.
Now, if you'll excuse me, I need to go and fire up Cursor... an idea has just hit me 💡