By Alex Moynihan • Published 24 Apr 2025 • 5 min Read

‍

How the rise of the citizen developer will transform how we create software

The Turning Point 2014: “Serial” drives the explosion of interest in narrative driven and interview style podcasts.

Early podcasters used RSS feeds (with enclosures) to distribute audio episodes, which Apple integrated into iTunes in 2005. While this made subscribing and syncing podcasts to iPods more accessible, it remained niche, mainly among tech-savvy users.

In October 2014, the true crime podcast Serial launched, becoming a cultural sensation, resulting in over 5 million downloads. The ecosystem subsequently exploded, driven by streaming services like Spotify & iHeart, smartphones that made access frictionless, celebrity entrants into the market, and maturing commercial models to monetise the medium.

In 2013, Global advertising sales totalled an estimated $45 million. In 2024, this was estimated at $3 Billion - $4 Billion. This makes Serial the Turning Point for podcasts and the wider broadcasting industry.

This was a democratisation of media and storytelling, giving anyone with a voice and a microphone the ability to reach a global audience, without needing a broadcasting license, publisher, or traditional media gatekeepers.

Similarly, Generative AI has enabled the democratisation of software creation, giving anyone with an idea the ability to reach a global audience, without needing diverse technology skills, a mature distribution network, or global partnerships. It excels in code generation due to its ability to leverage rules based and logic-driven structures, significantly enhancing prediction accuracy.

In late 2023, I used Anthropic and Open AI to generate the code required to complete an AI programming course. It was clunky but very effective, resolving runtime errors by pasting them back into the chat window to ask for solutions. At the time, platforms like Replit, Bubble, and Cursor were beginning to emerge as AI enabled coding tools that could support non-developers in building robust, scalable and market ready web-based applications. These tools evolved and scaled so quickly in the intervening period that this week OpenAI set its sights on Windsurf, another rapidly scaling AI coding company whose Annual Recurring Revenue surged from $40 million to $100 million in just months. According to Bloomberg, OpenAI extended a $3 billion acquisition offer.

To demonstrate the step change in evolution, I recently used Replit to build a fully functional app to manage all my priority activities based on user input, Kanban boards, emails and any other API enabled sources. The app prioritises my activities for the day based on how I like to manage them, their urgency, contact, due date, status, priority, activity detail and other characteristics. It learns from lagging or overdue actions, ignored actions, important contacts, why actions are cancelled and provides advanced analytics to help structure my time, providing recommendations. Adding new features is as straight forward as engaging in the chat window and helping your AI buddy tailor the features to meet your requirements. To build the fully functional MVP of the app took about one business day of effort from requirements to operations, for one person.

As a default, Replit used the following services for my app but can be configured to any organisational enterprise and technology architecture:

• Database: PostgreSQL with Drizzle
• Frontend: React with Tailwind CSS
• State Management: React Query
• Form Handling: React Hook Form with Zod
• UI Components: Shadcn UI

For the record I don’t even know what a React Hook Form with Zod is, not to mention Drizzle ORM.

While jaw-dropping, it’s not just the advances themselves, but the pace with which these platforms are proliferating and gaining traction. Extrapolating the rate of evolution we have seen in the last two years it is easy to conclude that we can expect exponential improvement in the next two years. What I did today will pale in comparison to we will do in 12, or 24 months. While nothing is certain, two immediate disruptions seem very likely from this accelerating evolution:

Disruption #1

Consumer to Consumer developers will create a viable marketplace, carving out a material slice of the enterprise software market

Individuals will, and already are, by virtue of our own subject matter expertise and ideas, the creators of software ready for consumption by other individuals or enterprises.

Niche markets will emerge and grow of like-minded consumers who derive value from applications and will not need the scale of a global technology provider to monetise and profit from these creations. This will evolve into groups of creators collaborating and creating apps that enhance their propositions as markets mature, taking on established enterprises.

Furthermore, large, global enterprises may face what Clayton Christensen coined the innovators dilemma, that outlines how large incumbent companies lose market share by providing what appear to be the highest-value products. However, new providers that serve low-value customers with poorly architected technology can improve that technology incrementally until it is good enough to quickly take market share from established business, which in turn accelerates the innovation cycle.

Three factors will influence the speed and traction of the C2C movement:

• A seamless and engaging distribution channel like Spotify is to Podcasts. Huggingface, Open AI, Azure Market place and others will need to find a way to engage a mass market in much the same way
• A Serial moment where a C2C app breaks through the niche barrier triggering a Turning Point for C2C app development
• The C2C market with methods to inspire trust without adding prohibitive burden on creators

Disruption #2

Upending the Software Delivery Lifecycle and making the Business Analyst (BA) the centre of the enterprise in software creation

The BA has traditionally played the role of interpreter between business and technology. Whatever the enterprise methodology (Agile, Waterfall or DevOps), the BA has been central only in the early and final stages of delivery. With AI assisted development, this is upended, enabling the BA to be the author of software creations end-to-end, with technology being the editor and publisher of these creations.

Industrialising this approach requires the right approach to re-designing the technology delivery operating model. The roles of business stakeholders, BA and technology stakeholders will change significantly, with some of the key SDLC stages possibly evolving to include:

‍

‍

In a Consumer-to-Consumer scenario, there is not the rigour required to make available rapidly developed software, but in industry enterprises such as Banking and Financial Services, with a high degree of regulatory scrutiny, a more formal program of transition will be required to avoid a Shadow IT or proliferation of the new Excel Macro, but the acceleration to value will be immediate and very impactful. This is further accelerated by AI enabled transformation of the IT function with Scan AI.