#50 - Is AI A Faster Horse or the Next Tesla?

How Generative AI is Transforming Technology, Business Models, and Our Approach to Problem-Solving

The dilemma

AI in general, and GenAI specifically, are common trends and buzzwords. Recently, I was involved in several discussions about how to treat them in our roadmap. For example, a theme like “Add AI to the product” is poor as it discusses the solution without any context regarding the problem. Alternatively, using AI to solve a theme like “Improve retention” could make it seem as though AI is treated as a side project.

So, what should we do? Should we take the problems we identified so far and see where AI might help? Or should we focus on AI and discover whatwe can do with it?

To address this question, I drew on Henry Ford’s analogy.

If I had asked people what they wanted, they would have said faster horses.

AI, in this case, is like building a car—a transformative technology that changes the market entirely. However, I’m not a fan of clichés. Instead of stopping here, I decided to look at other technology trends to understand how they fit into this perspective and how to resolve my conflict.

Technial references

Microservices vs. Microfrontend

Microservices are something I see as disruptive technology. In the old days, everything was monolithic. Software products had one codebase. All the code was in one repository, and everyone contributed to that codebase. Eventually, that codebase would turn into the final product for the customer.

Microservices approached the complexity and maintenance challenges of monoliths in a completely different way. They abstracted the dependency and collaboration challenges by allowing the creation of multiple services, each focused on a specific context with limited ownership, while communicating with other components through defined interfaces.

This solved the problem but created an entirely new market. Suddenly, companies had tens or even hundreds of microservices to maintain. This shift required new infrastructure, new ways for services to communicate with one another, and new skills for developers. Previously, developers didn’t need to deploy or create new services—those were already part of the monolithic system. Now, developers had to manage things like service-level agreements (SLAs) and figure out how services should communicate effectively, such as through REST APIs.

Micro frontends, on the other hand, are more of an optimization than a disruption. They take the principles of what microservices achieved on the backend and apply them to the frontend. With this approach, codebases are simpler, and teams are more independent in deploying and utilizing the tech stack they want.

However, the challenges remain largely the same. While microfrontends solve similar problems slightly better, they don’t open up entirely new markets or business models the way microservices did. Instead, they represent an incremental improvement in managing complexity rather than a foundational shift.

Internet and cloud computing

Looking in a different direction, the Internet itself was arguably one of the most disruptive technologies ever, opening up an entirely new world of possibilities. However, advancements like 4G and 5G, while impressive, are incremental improvements. They enhance the existing infrastructure by providing faster Internet and greater accessibility, but they don’t change the market.

Cloud computing, on the other hand, represents a critical shift that reshaped the industry. Before its introduction, launching an online business or offering software came with the significant overhead of purchasing and maintaining infrastructure, such as servers and physical machines. This was a major blocker for many businesses.

Cloud computing removed this problem entirely, introducing a new way of working. In just a few minutes, you can now have a virtual server hosting your solution anywhere in the world. This transformation reduced entry barriers and created a new type of business.

How to identify a disruptive technology

When considering AI, and generative AI specifically, we need to evaluate just how disruptive it will be. Is it solving existing problems better, much like Henry Ford’s notion of “faster horses,” or is it creating entirely new markets and business models, maybe even a “Tesla”?

To answer this, I propose assessing AI through the following factors:

1. New Markets and models: Does AI create entirely new markets, product lines, or business models within its segment?

2. Surface New Problems: Are there problems we hadn’t previously considered that AI now enables us to solve?

3. Obsolete Problems: Are there issues we currently face that AI will render irrelevant altogether?

These questions are important, as they can help us shape how we integrate them into our strategies and roadmaps.

My answers

I’ll admit to having a bit of imposter syndrome here, but my honest opinion is “yes” to all three questions.

First, there are undeniably new markets emerging: AI agents, generative AI platforms, and AI-driven ecosystems that didn’t exist before. These developments have also created new opportunities, such as the growing demand for faster and more capable processors and the significant rise in GPU usage.

For the second question regarding the new problems AI surfaces, it brought up privacy, intellectual property, and ethical concerns, to say the least. Over time, I believe these areas will have their own models and businesses.

As for the third question, obsoleting problems, definitely. I’ve experienced it firsthand. When I worked on a side project, I used to struggle with deploying it. Tools like Replit now abstract away the complexities. Similarly, in my past work on automations, I often faced challenges in finding the right steps and setups to achieve my goals. Now, with just a simple prompt, the solution is built right in front of me. I no longer need to be as knowledgeable or spend as much time figuring it out as I did before.

My conclusions

So, what does it all mean? In my opinion, it means that we need to view AI, and generative AI specifically, as a strategic opportunity we must position ourselves against. It’s not just another technology to solve existing problems; it’s a fundamentally new way of thinking that brings with it an entirely new type of problem-solving. To truly succeed, we need to consider not just the problems of today but also the problems of tomorrow—and the solutions of the day after.

For example, consider the context of e-commerce. It’s not far-fetched to imagine a future where my smart home knows my refrigerator is nearing the end of its life. It could proactively send me a suggestion for a replacement, and with a single click, the purchase is completed, and the new refrigerator is delivered and installed a few days later. This is a problem that doesn’t exist today but very well might tomorrow. Where would this position physical stores? Search engines? Technicians?

Another example comes from the travel industry. Today, the process of finding the cheapest flight ticket sustains an entire business model. However, in the future, AI agents that bid, crawl, and negotiate on my behalf could eliminate that market entirely. Companies and products must evolve to meet this new way of thinking.

I’m not suggesting that AI is the answer to everything. It certainly has its downsides and limitations, many of which we’ve yet to fully understand. But that doesn’t change the fact that it’s here to stay. It’s a disruptive technology that will revolutionize our lives, and to stay ahead, we must ride the wave and adapt.