AI Insights

For years, software delivery has followed a fairly simple pattern – if demand increased, you hired more engineers. If programmes got bigger, teams got bigger, and value largely scaled with headcount. That model made sense when building software was hard, slow and expensive, and engineering capacity was the constraint. But AI is breaking the old model apart in some fundamental ways, and that impacts not only our customers at Godel, but also the products and services we offer them.

Large parts of software engineering aren’t scarce skills anymore. Code generation, testing, documentation and analysis are no longer the things that limit progress. Writing software is becoming easier thanks to AI, but understanding what you should build, how you should build it, and why it matters is where the value is now delivered. Creating the software is increasingly the easy bit.

Internally at Godel, that forced us to be honest with ourselves. If we carried on delivering in the same way, we’d be optimising a model that is being steadily commoditised. AI allows much smaller teams to deliver the same output as larger ones, and in many cases, the expectation isn’t just parity – it’s significantly more. You don’t need fifty people doing what five can now do, and you certainly can’t build a long-term business assuming customers will continue to demand that. We’ve adapted to those rapid changes.

This shift isn’t just theoretical either– it’s already visible in customer behaviour. In a recent survey, 67% of our existing customers told us they’re developing AI into their products now, not as experiments, but as real capability. Boards, investors, customers and competitors are all applying pressure, and that pressure lands firmly on the desk of the CTO.

At the same time, that pressure is now extending into delivery itself. 90% of those organisations expect to be using AI within their delivery lifecycle in 2026, which means the way software is built, tested and released has to change alongside the products. CTOs are being pulled in two directions at once, delivering new capability while still producing reliable software every week.

AI doesn’t reduce demand either, it increases it. Cheaper and faster delivery leads to ambition and growth, which in turn creates more complexity and decision-making. Customers are asking us how to navigate this without breaking their products, or their credibility.

Buy vs build

The other driver for change is that AI breaks some of the assumptions we’ve relied on for years, particularly around buying versus building software. For over a decade, the advice was buy, don’t build. SaaS won because it was faster, but AI disrupts that logic. The advantage of off-the-shelf software no longer comes from speed alone, and when differentiation increasingly lives in data, workflows and decision-making, buying doesn’t guarantee an edge either. That doesn’t mean everything should suddenly be built, but it does mean the old shortcuts no longer protect you.

At the same time, delivery models themselves are under strain as inefficiency is exposed. Large, slow, people-heavy programmes are becoming harder to justify. Investors don’t care how many people you employ, they care how quickly you can move and what outcomes you can prove. Most CTOs are caught in the middle, with pressure from above and teams adapting below. Decisions get made later, under more pressure, with fewer real options.

How Godel’s changed

How we deliver. We now build with AI embedded into our engineering process, not as a bolt-on, but as part of how work gets done. Smaller teams, faster feedback and delivery models designed for AI-augmented engineering, not scaled headcount.

What we help customers do. We’re not just adding AI features, we’re helping organisations move from where they are today towards AI-native ways of working, when it makes sense for their business.

How we think about value. This isn’t about selling people or hours, it’s about solving real problems quickly and with clarity, in a market that has become noisy and confused.

The biggest risk right now for CTOs isn’t choosing the wrong AI tool, it’s sticking with delivery and buying models that no longer fit the reality AI has created. Godel changed because the rules have changed. Those who ignore the implications will find themselves reacting later, when the cost is higher and the room to manoeuvre is much smaller.

Joe Wolski, CTO, Godel

Posted 18 Mar 2026

share