For decades, the viability of a software project was constrained by one fundamental question:
Can we build it?
Ideas were plentiful. Business problems were abundant. But translating concepts into working software required teams of analysts, architects, developers, testers and project managers. The cost, time and complexity of software development acted as a natural filter. Many ideas never progressed beyond PowerPoint presentations and whiteboards because the economics simply did not justify the investment.
Artificial intelligence is beginning to remove that constraint.
Over the past months, I have spent considerable time experimenting with AI-assisted software development. As a digital consultant in the banking industry, I have built applications such as the BICon Investment Decision Lab, which supports portfolio construction, investment proposals, benchmark design, risk assumptions, governance and documentation. I have also developed a digital transformation steering platform covering digital strategy, AI governance and regulatory – EU AI Act – compliance.
The End of "Can We Build It?"
The most interesting observation has not been the speed of development. It has been the changing nature of the decision-making process.
The question is no longer:
Can we build it?
The question is increasingly:
Should we build it?
This distinction may sound subtle, but it represents a profound shift.
When Implementation Becomes a Commodity
As software creation becomes increasingly commoditised, the limiting factors move elsewhere. Coding capacity becomes less important. Competitive advantage shifts towards ideas, domain expertise, intellectual property, proprietary data and distribution.
At the same time, the market is likely to be flooded with applications built on half-formed ideas but implemented with remarkable technical quality. When implementation becomes a commodity, technical execution alone ceases to be a meaningful differentiator. The ability to identify important problems, formulate compelling solutions and reach customers effectively becomes far more valuable.
Many concepts that would previously have failed a business case can suddenly become economically viable.
The Rise of the Maximum Viable Product
For years, the software and startup ecosystem has embraced the concept of the Minimum Viable Product – building the smallest possible solution capable of validating a market need. That philosophy emerged because development resources were scarce and expensive.
AI may be pushing us towards a different paradigm.
As the cost of creating software falls dramatically, organisations can increasingly afford to build richer and more comprehensive products from the outset. Functionality that would once have been dismissed as unnecessary, excessive or unaffordable can now be implemented with relatively little additional effort.
One could describe this as the emergence of the Maximum Viable Product.
Instead of asking what functionality can be removed to make a project economically feasible, organisations can increasingly ask what additional functionality could create value for users. The economics of software are changing from scarcity to abundance.
The New Constraint: Imagination
The economics of software are changing from scarcity to abundance.
The constraint is no longer development capacity.
The constraint is imagination.
In a world where almost anything can be built, the ability to identify opportunities becomes more valuable than the ability to implement them.
Why Banking Could Be Disrupted
This shift has important implications for banking and financial services.
Historically, large financial institutions enjoyed significant advantages because they could afford large technology departments and proprietary systems. Smaller banks, wealth managers and advisory firms often struggled to match the functionality available to larger competitors.
AI-assisted software development may begin to erode that advantage.
If experienced investment professionals can directly create sophisticated portfolio construction tools, reporting systems, advisory platforms, governance solutions and analytical applications, the source of competitive differentiation changes.
The New Sources of Competitive Advantage
Software development will not disappear. Rather, it may cease to be the primary source of scarcity.
Success may increasingly depend on:
- Intellectual property
- Domain expertise
- Proprietary data
- Investment processes
- Customer relationships
- Distribution
These factors are considerably harder to replicate than software itself.
Knowing What Should Be Built
The coming disruption may not be that AI replaces software developers.
It may be that software development itself becomes a widely accessible commodity.
The winners may not be those who can build.
They may be those who know what should be built.
That is a very different world from the one the software industry has operated in for the past three decades.
And we are only beginning to understand its implications.
For those interested in exploring this idea further, I have written several articles on the topic at BICon.li and published a number of experimental AI-assisted applications, including the Investment Decision Lab. Apart from the AI Act platform, these projects are not commercial products but practical experiments designed to explore what becomes possible when software development is no longer the primary constraint.