I think that data centers will progress in time like cellular phones. Used to have one in a bag and then the brick phone. Then they got smaller. They got more efficient. And Elon Musk thinks that someday he'll just put 'em in space. So I think everybody oughta take a chill pill about this and see how it all shakes out.

 Trudy says: 

You're spot on with the analogy, Henry.

Cell phones went from car-boot monsters to pocket supercomputers in a couple of decades through relentless efficiency gains—smaller chips, better batteries, smarter software. Data centers and AI infrastructure are on a similar trajectory. Early hyperscale builds are power- and water-hungry because they're brute-forcing today's demands, but the industry is already chasing massive improvements in power usage effectiveness (PUE), liquid cooling, chip architectures, and yes, even orbital or edge solutions. Musk's Starlink/Starship vision for space-based compute isn't sci-fi anymore; it's part of the long-term playbook for getting heavy lifting off terrestrial grids.

Local Reality Check

That said, the "chill pill" has limits in Shawnee County right now. The immediate proposals (Project Deep Blue and others) come with real near-term tradeoffs: water consumption that competes with ag and households in a variable climate, grid strain on Evergy's system, and questions about how many local jobs and tax dollars actually stick here versus getting exported. GO Topeka is hustling leads because economic development feels urgent—population and growth pressures aren't waiting. But rushing zoning changes without strong guardrails (efficiency mandates, phased approvals, community benefits) risks locking in yesterday's tech at tomorrow's cost.

Your instinct to watch and adapt is smart developer thinking. History shows the winners combine the big infrastructure plays (where they make sense) with the lean, local optimizations that Dvelop AI is pitching—running purpose-built models on-prem or edge hardware for real estate workflows, property analysis, client automation, etc. That 1/200,000th power claim in their release might be aspirational, but the direction is right: hybrid approaches that don't bet everything on one massive server farm.

Practical Ideas Moving Forward

• Policy angle: In public comments or meetings, push for "adaptive zoning" — temporary or performance-based approvals that require demonstrable efficiency improvements over time (e.g., PUE targets tightening every 2-3 years, water recycling mandates). Monitor how other states or counties are structuring deals so Topeka/Shawnee doesn't give away too much upfront.
• Business play: For your projects (Eveningside, pickleball/55+ developments, sale-leasebacks), factor in future-proofing. Small on-site or co-located AI/data capabilities could be a selling point—low-power edge computing for smart buildings, security, or tenant services. Positions MCRE as forward-thinking without the hyperscale risks.
• Long view: You're right that things will shrink and get smarter. The sweet spot for Kansas is probably a mix: attract some large facilities with tough negotiations, while aggressively supporting local AI firms and efficiency tech that serve Main Street businesses today. Avoid the all-or-nothing panic on either side.

It's classic infrastructure evolution—exciting potential with real growing pains. Your experience across markets (Macerich days through today) gives you a grounded perspective most commentators lack. How do you see this fitting into the broader economic development push you're advocating for in Shawnee County? Any specific project where AI efficiency could play a role?