Yuri Gubin is Chief Technology Officer at DataArt, working with organizations integrating AI into complex engineering environments.
getty
One analogy I keep coming back to is owning a Ferrari just to pick up your kids from school. Having a faster car does not suddenly allow you to ignore traffic lights, speed limits or the realities of the road. The bottleneck was never the car itself. In many ways, that is what is happening with AI adoption inside organizations today.
For years, software development was considered the constraint. Businesses would come up with ideas, define requirements, align stakeholders and then wait for engineering teams to build the solution. Now the equation is changing. With AI-assisted and agentic development, engineering organizations can produce prototypes, features and even entire solutions much faster than before, so the bottleneck is moving somewhere else.
What I increasingly see is that organizations are struggling to keep up with their own engineering capacity. Development is accelerating, but the rest of the organization often still operates at the same pace as before.
You can already see this in practical situations. Years ago, preparing an ad hoc business or operational report could take weeks. Teams needed analysts, researchers, coordination and multiple iterations before leadership could even review the findings. Today, with sufficient governance and data maturity, AI can generate meaningful operational insights in less than an hour. The question becomes whether the people responsible for making decisions have enough time and capacity to process that information.
That is why I do not think the story is simply that AI is replacing developers. The larger issue is that many organizations are not yet prepared to consume the increased productivity AI creates. They cannot produce new ideas quickly enough. They cannot validate opportunities quickly enough. They cannot adapt decision-making processes quickly enough, and this is where I believe the next major opportunity exists.
One of the first places where this becomes visible is product ideation, because before anything gets developed, somebody still needs to define what should be built and why it matters. What I increasingly see is AI becoming a sparring partner for product teams, helping them challenge assumptions, pressure-test concepts and refine ideas much faster.
A product team can use AI to validate hypotheses, challenge concepts or simply think through ideas during brainstorming. There is a big difference between talking about an idea in theory and putting a prototype together in a few hours so people can react to it. Once somebody can click through something, see it and react to it, the discussion changes completely and the feedback becomes much more practical.
Another capability comes from processing unstructured information, since companies already sit on enormous amounts of customer feedback, reviews, research, operational data and industry commentary, but historically, going through all of it manually took far too much time. AI can now process that information much faster, helping teams understand where opportunities or problems may exist.
The same thing applies to market analysis, where companies constantly try to understand where the market is moving, how customers think, what competitors are doing and whether demand is changing. Public companies publish enormous amounts of operational and financial information every quarter and historically, somebody had to spend hours reviewing reports, synthesizing findings and trying to connect the dots.
AI can monitor reporting continuously, analyze sentiment and help generate hypotheses much faster, allowing teams to spend less time collecting information manually and more time interpreting what matters and deciding what to do with it.
Customer feedback is also changing because the old model was often static and impersonal, and surveys or forms could only get you so far. With AI-driven interactions and conversational agents, companies can gather feedback earlier and in ways that feel much closer to a real discussion, helping teams understand reactions faster, identify friction points sooner, rethink priorities and iterate on products much more quickly.
Decision making is changing as well, though I do not think AI should replace executives or product leaders in strategic decisions. AI can help people evaluate options much faster. For example, a SaaS company with a free tier can model different pricing structures or conversion assumptions using its operational and financial data, allowing leadership teams to test scenarios, challenge hypotheses and better understand possible outcomes rather than relying solely on static reports or assumptions.
I use similar approaches in my own work when evaluating potential service offerings. Sometimes we identify a solution that works well for a specific type of client, but then the question becomes whether the same idea applies somewhere else. I use AI to challenge those assumptions and stress test the business case. For example, I may ask whether the same service would solve meaningful problems for nonprofits, healthcare startups or an entirely different market segment. Sometimes the exercise supports the idea, and sometimes it exposes limitations or gaps I had not considered before.
I think this is the broader point organizations need to understand about AI adoption. AI adoption is not only about helping developers write code faster or automating routine tasks. Software development is changing rapidly, but engineering alone does not create successful products or successful businesses. Every layer of the organization now has opportunities to improve how it thinks, analyzes, validates and makes decisions. Companies that learn how to apply AI across the entire operational chain will be much better prepared to benefit from the increased speed modern engineering teams can already deliver.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。