MB533: The AI Revolution Is Here (Will Investors Be Left Behind?)

Show Notes

Jake Heller is third-generation from a private homebuilding family that built several thousand homes across Southern California, Arizona, and Nevada.

In August 2025, he launched the AI for CRE Collective, a members-only platform that tests AI tools, publishes workflows, and tracks what's actually working in commercial real estate. The community has grown to over 800 members and now operates as one of the leading AI enablement platforms in the industry.

The episode closes with a concrete starting point: Claude in Excel. Jake calls it the conversion event for people who have been resistant. Drag an offering memorandum into Claude, tell it to extract the rent roll and P&L into your spreadsheet, and watch it work. You don't need to be technical. You don't need a developer. You need a $20-per-month account and a willingness to run one experiment. That's the assignment.

Key Takeaways

AI multiplies your fundamentals

If you have a solid understanding of how to underwrite a deal, research a market, and evaluate risk, AI can make you significantly more productive across all of those tasks.
If your fundamentals are weak, AI produces confident-looking output built on bad inputs.
Don't skip the foundational work and treat AI as a substitute.

Context is what separates useful AI output from hallucinations

When AI doesn't have enough context, it makes assumptions.
The job of an orchestrator is to minimize the number of assumptions AI needs to make by providing precise context upfront.
Explain your underwriting criteria, your market assumptions, your return targets, and your model structure to AI.
Over time, a well-trained prompt library and accumulated context produce much cleaner output than cold prompts do.

Measure ROI before committing to any AI workflow

Establish a baseline first. If underwriting one deal takes you one hour without AI, that's your number.
Then factor in the full AI-assisted time: setup, prompt iteration, reviewing output, making corrections.

Don't give AI authority over decisions or legal documents on serious transactions

AI is useful for modeling scenarios.
Making the decision is still your job: never outsource judgment to AI on a transaction that involves millions of dollars.
Relationships, market knowledge, and deal instincts built over years don't transfer to a language model.
If being wrong has significant financial or legal consequences, a human with domain expertise needs to be in the loop.