Most real estate content tracks the market. We track the execution. Every Saturday, get the specific deal structures, underwriting frameworks, and capital strategies we are using to navigate the current cycle.
The Printing Press Moment (And How to Actually Use It)
|
Boris Cherny, Head of Claude Code at Anthropic, introduced me to the printing press analogy. Before the printing press, books were scarce. Hand-copied by scribes in monasteries. Expensive. Accessible only to elites. Johannes Gutenberg's printing press (1440) changed everything. MORE books. CHEAPER books. Mass literacy. The democratization of knowledge. But the transition was brutal. Scribes lost their livelihoods. Monasteries disrupted. Religious institutions that controlled information faced an existential crisis. The scribes saw only destruction. Society saw the printing press. We're the scribes right now. February 2026: CBRE down 12%, JLL down 20%, office REITs down 6.7%. Capital markets pricing in the displacement phase. But the abundance phase is coming. The question: Position during the transition or get displaced by it? The Printing Press Moment Is Here (And It's About Agentic AI) This isn't about basic AI. It's about agentic AI, AI that takes action autonomously. Claude Code, the AI agent that codes and writes software for you, just celebrated its one-year anniversary on February 21, 2026. One year ago, this didn't exist. In that same window, OpenClaw (originally "Clawdbot") went from a weekend project by Austrian developer Peter Steinberger to 149,000 GitHub stars. OpenAI hired Steinberger in February 2026 to drive "the next generation of personal agents." Agentic AI executes tasks autonomously: writing code, managing email, controlling calendars, navigating applications, making decisions without waiting for you. CRE is way behind this curve. Multifamily hit 99% adoption, but that was basic AI (chatbots, dynamic pricing). The agentic AI wave just started. February 2026. Claude Code Hackathon. 500 Builders. One Week. First place: Mike Brown. California personal injury attorney. Not a software engineer. Built CrossBeam, a permit compliance tool that speeds up California's permitting process. Third place: Michal Nedoszytko. Cardiologist in Brussels. Built postvisit.ai. Turns visit transcripts into personalized health guidance. His first user: his own confused patients. A lawyer and a doctor. Building software applications. In one week. Boris Cherny: "At this point it's safe to say coding is largely solved. At least for the kinds of programming that I do, it's just a solved problem." That's the printing press moment. Prior to the printing press, written knowledge was only accessible by the few. Coding has now reached that arc. It is now accessible to all because AI codes for you in the same way printing no longer occurs by hand. This is what Anthropic expects: Anyone can code, create software, or tailor software to their needs. Not just Boris and his world-class engineering team completing 49 separate software updates in two days (work that would normally take weeks). A lawyer navigating building codes. A cardiologist treating hearts. This is capability democratization. The printing press for software. Two AI Paths, One Uncertain Outcome Anthropic/Claude: Workflow Augmentation AI becomes your institutional-grade intern. You direct, it executes. Underwriting, portfolio analysis, IC prep, asset management, all democratized. Andrej Karpathy (ex-Tesla AI Director, OpenAI founding member): "Telling AI what to write is like directing an intern. Once you control software at scale, you can never go back." This path says: Judgment stays valuable. Experienced operators with AI leverage dominate. Google/Gemini: Raw Intelligence Scaling Gemini 3 achieves PhD-level reasoning. "Deep Think" mode solves problems that challenge advanced AI systems. The goal: artificial general intelligence that surpasses human cognition. This path says: Raw intelligence wins. Humans become strategists or obsolete. We don't know which path wins. Gary Marcus: "You're not getting a machine that is Jeff Dean in the next decade." Maybe human judgment can't be replaced. Or Daniel Kokotajlo: "AGI by ~2030 or bust." What we know: The transition is now (2024 to 2028). It's a skills-based function. You Can't Read Your Way to Agentic AI Competence You must use it daily. But first, the mindset shift. There is a learning curve and you must take the time to learn the tools. It's more of a mindset. Instead of using AI agents like an answering machine, you need to treat it like an employee, and a capable one at that. To this end, you've got to learn to see what's possible. Tell it what you want to achieve (the hardest part) and brainstorm with it to plan how to approach the task or problem. Go back and forth with it, then tell it to execute the plan. The AI agent can figure out the resources and tools it needs and will execute it with your permission. How I'm Using Agentic AI Right Now 1. Claude Desktop App for Excel Model Comparison Comparing two different capital stack models, identifying variance, implementing changes to the new version. Would have taken hours. Claude Desktop App did it in minutes. 2. Claude Desktop App & Gemini for Browser Control Both Claude Desktop App and Gemini can take over my Chrome browser to execute searches, research, and execute manual work, not just on the web, but within web-based apps. Work that would have taken weeks, now done in hours. 3. Gemini Deep Research for Source File Generation Automatically generating source files, synthesizing research, and generating research reports. Calculating analysis grounded in research. The Pattern: I direct. AI executes. Three Practical Steps 1. Pick One Agentic Tool, One Workflow
Pick one 2-plus hour per week manual workflow: underwriting analysis, variance review, market research, IC memo drafting. 2. Judgment Required: Your Professional Advantage Your advantage as a professional is you can provide skilled and actionable direction. Bad: "Analyze this acquisition." Good: "I'm evaluating a 120-key extended stay hotel acquisition in Austin, TX near the Domain. Purchase price $18M. T-12 NOI $1.45M. I'm considering 75% LTV bank debt at 7.5% and 25% LP equity. Help me build a scenario analysis framework. First, outline the key variables we should stress test (occupancy, ADR, OpEx, CapEx, exit cap rate). Then create a decision matrix showing which scenarios justify proceeding at different return thresholds. Once we agree on the framework, I'll have you build the model." Specificity. Context. Collaborative planning. Then execution. 3. Validate Everything AI hallucinates. Veterans have the advantage, you know what "right" looks like. Practical validation tips:
Rule: If the output drives a capital decision, validate before relying on it. Test on Your Capital First GA 2-Pack is my laboratory. I test every AI workflow on my own operations before advising clients. Build hybrid teams. Veterans direct and validate. Young talent implements and scales. The Abundance Phase The printing press didn't destroy writing. It created MORE writers. Agentic AI will create:
But only for operators who survive the transition. The transition is uncomfortable. The deflationary pressure is real. The uncertainty is legitimate. But the abundance is coming. Position now. -Damon P.S. Reply with one workflow you are implementing or if you want to talk through your capitalization or portfolio growth plans. You can Schedule a call or message me on LinkedIn. I get deals done and can help you grow. |
The Eternal Edge
Most real estate content tracks the market. We track the execution. Every Saturday, get the specific deal structures, underwriting frameworks, and capital strategies we are using to navigate the current cycle.