April 2026
The Real Boss™ is a very gifted photographer by trade. When the iPhone came along, suddenly everyone thought they were a photographer. They didn’t need to buy fancy cameras anymore, just use the phone they already owned. And no need to pay a pro for those holiday pics!
I’ve seen some variation of this a thousand times: a mom takes a picture of the volleyball team with her iPhone. The Real Boss™ takes a picture with her iPhone. They compare. The mom is stunned by how much better Sarah’s pics are. “What iPhone do you have…”
The barrier to entry to take photos has come way down…but that doesn’t turn everyone into a photographer. Talent, combined with years of experience and lessons, translates into great photos. A $10k camera in my hands is basically a waste of $10k because I haven’t learned those skills.
I was at an event in Dallas this week and a lunch speaker cancelled at the last second. To fill the time, I was asked if I could demo how I use Claude. At some point I was asked by an attendee, “Do you worry this will make LoanBoss obsolete?” He asked it much more directly, but with wording that your email filters would likely block.
“I have thought about that every minute of every day since the first time I tried ChatGPT!” The room laughed nervously, pitying me.
It’s a fair question. AI makes software feel buildable. Screens feel cheaper. Workflows feel easier to replicate. But there is a huge difference between building software and building software that survives real-world use, edge cases, bad inputs, and the thousand tiny failures that show up only after years in production.
People think they pay SaaS companies for the screens. They don’t. They pay for the hard-fought lessons, the expertise, the wisdom gained from battles won and lost. They pay for the team behind it, so their own team can focus on what matters most.
Isn’t that the same pitch you make to investors? Every single investor could buy real estate on their own, but does that make them a real estate expert? Aren’t they better off trusting you? Why is technology different?
For years, many companies avoided paying for software for a very familiar reason: “Our current patchwork is not perfect, but it is close enough. Yes, it is inefficient. Yes, it has friction. Yes, it is dependent on a few employees who know where everything lives. But it basically works, and the unknowns of a new vendor feel riskier and more expensive than the known shortcomings of the current mess.”
AI has not eliminated that logic. It has just given it a more exciting wrapper.
Instead of saying, “We don’t need outside software because our spreadsheets are fine,” people now say, “We don’t need outside software because we can build our own with AI.”
Are you really going to design property accounting software better than Yardi? Better than a company that has spent years absorbing customer pain, strange exceptions, compliance headaches, and all the ugly realities of production software?
If our product was basically a set of screens on top of obvious workflows, then yes, I would be more nervous. AI is going to make thin software feel very thin. A lot of software was more replaceable than anyone wanted to admit.
But if the product contains accumulated judgment, domain expertise, process design, data structure, and the embedded lessons learned from solving the same messy problem thousands of times, AI does not erase that value. It magnifies it.
In fact, I would argue the real opportunity of AI comes after the system is built, not before.
The flashy part is using AI to create software. The valuable part is what happens once you have the right data, structured the right way, inside the right system. That is when AI gets dangerous in a good way. That is when it can analyze, recommend, surface patterns, automate exceptions, and drive better decisions.
But AI is only as useful as the foundation underneath it.
If the underlying system is sloppy, fragmented, poorly designed, and full of inconsistent data, AI just helps you move faster in the wrong direction.
Nearly every real estate firm I’ve spoken with over the years has tried or at least considered building custom software. What makes AI dangerous is the appearance that they can suddenly realize that dream. Just because you have an iPhone doesn’t mean you’re a photographer. Just because I’m able to view a listing online doesn’t mean I should compete with you on real estate investing. Whatever I might save on fees is likely to be more than offset by a variety of other downside risks.
Some of you will ignore this and try anyway. “Yardi is different because it’s already in place. But I can build a loan management system from scratch.” I get it. I would probably be one of you. But I’ve learned a lot of painful (and expensive) lessons along the way, and now I am a firm believer in outsourcing to the experts. It doesn’t take much trial and error to exceed the cost of a SaaS provider.
The screens were never the whole value. The lessons behind them were.