Spring’s housing market has been a snooze—except for San Francisco, where AI money is rewriting the rules.

Luxury home sales in the backyard of Anthropic, OpenAI, and other startups have soared 48% year over year. Prices are up nearly 10% annually to a median of $6.7 million, the highest in the country. One six-bedroom in Cow Hollow that was listed for $7.95 million sold for $15 million—88.68% over ask, a record not seen this century. Some home sellers, rather than settle for double their asking price, are fishing for Anthropic or OpenAI stock.

“People in AI earning $40,000 to $70,000 per month have been renting but are ready to buy,” says Alan Mark, a San Francisco–based real estate strategist and consultant. Investor Jason Nesbitt at Peachtree Homes agrees, adding: “In the past, buyers were primarily senior executives. This time, younger AI workers already have high salaries and equity upside. They’re house hunting earlier.”

These purchases are more than a place to live. “Many who have doubled or tripled their net worth are looking to redeploy concentrated positions in technology equities into hard assets,” says Simona Martin, a real estate agent with Christie’s International. “They understand that equities can be volatile, whereas real estate has durability as both a lifestyle asset and a store of value.”

How to get in on the AI housing boom

If you’re lucky enough to own property in San Francisco already, it may be tempting to sell and cash in. But wait. “If you have it, hold onto it,” says Mark, because the AI homebuying spree has just begun. “Once the IPOs happen, there’ll be an even bigger flood of capital.” In one of Mark’s two-bedroom rentals, a tenant paying $5,000 per month moved out. The up-to-date market rate? $8,500.

If you’re hoping to buy without a bidding war, Mark suggests venturing to up-and-coming areas within easy commuting distance of AI epicenters like Jackson Square or Telegraph Hill. Nesbitt has gotten “geographically picky,” narrowing his search to neighborhoods rich in founder/VC networks, such as Pacific Heights, Presidio Heights, and Noe Valley.

Even so, “We typically don’t find ‘cheap’ properties in the conventional sense,” Nesbitt says. Keep in mind: If a house needs work, AI buyers typically won’t touch it—and this is where money can be made. In June, Nesbitt bought a fixer-upper for $625,000, then spent $150,000 on tricking it out with a “tech campus” aesthetic: an open layout, wide-plank flooring, a dedicated office, and smart home features throughout. The turnkey home attracted 15 offers, driving the price up over the $1.2m ask to $1.5m.

The AI housing boom may have started in San Francisco, but is spilling into other markets. Nationwide, the median price of luxury homes (in the top 5% for their market) rose 3.6% year over year to $1.39 million, more than double the gains for lower-priced properties. In hot spots like New York City and Miami, multimillion-dollar homes have even been listed for (you guessed it) Anthropic stock.

Landscapers love these five backyard trends—but are ditching these four.

The state with the highest incomes isn’t California or New York. It’s a place you’ll never guess.

🛒 The list of America’s “starter markets” is down to just six major cities where prices remain under $300k.

Fires are spreading across rent-stabilized buildings—and the reason isn’t what you think.

A former Apprentice star is giving away a former college campus in Vermont. Fair warning: It drove him nuts.

Popular in Europe, imported to the US, Seattle’s first “social housing” complex has a waitlist over 10,000 long.

England has a problem: too many country homes. No wonder they’re getting price cuts.

The highest penthouse in Queens, NY, has a view of the Poconos—and an insane rent to match.

This couple has lived in a converted van for three years, with a dog. The benefits go way beyond saving money (think clearer skin).

Chickens and thongs and ivory! Oh my! You won’t believe what real estate agents have had to hide when showing a home.

Data centers may be the backbone of the AI boom, but no one wants one in their backyard. Jeff Gyzen at Arcadis, who has designed these buildings for decades, is trying to change that, starting with the Terra Ventures Data Center in San Jose, CA, which is expected to open in 2027. Read on to find out why everything you thought you knew about data centers is out of date.

Q: How did you get into this field? “I’ve been designing data centers for 43 years. My first was in 1983 at IBM. It was pretty much steady until AI. Five years ago, a typical cabinet ran about 8 to 12 kilowatts. Now we’re looking at racks hitting a megawatt. So your overall power draw has gone through the roof.”

Q: How has data center design evolved? “In the past, data centers just took from communities. Public opposition has started to change that. I like to say we’re trying to make data centers better neighbors. Energy-wise, the Terra Data Center I’m designing operates in ‘island mode’ where we generate all power on-site so we’re not pulling from local utilities at all. We’re using natural gas fuel cells, which run at about 90% efficiency, far more efficient than traditional combustion processes.”

Q: How else are data centers improving? “Historically, data centers were notoriously ugly big gray boxes. Terra is a genuinely beautiful building. And since its fuel cells create carbon dioxide, which can increase produce yields—especially strawberries and tomatoes—we’ll use that for vertical greenhouses and sell what we grow at a local community farmers market.”

Q: What’s your advice for people who might be buying property near a data center? “Proximity to a traditionally designed data center is probably not a great investment. But near a data center designed like Terra? It can actually improve property values, tax revenue, infrastructure, schools, streets. So there’s a real financial upside for property owners in data center markets. I wouldn’t buy across the street, but a couple of blocks away is no big deal, assuming the data center is considerate of the local community.”

Click here for more surprises about modern-day data centers.

This week, we spoke to Dan Beit-Or at Simply Do It about investing in Nashville, TN.

Average home price: $436,355 (down 3.2% YoY)
Homes that sell over list price: 10.7%
Homes that sell under list price: 69.9%
Average rent: $1,810/month

The pros: Thanks to Music Row, “everyone wants Nashville,” says Beit-Or. “The area also boasts strong population and job growth—hello In-N-Out Burger—and landlord-friendly laws.”

The cons: High costs and competition. “I don’t recommend buying Downtown,” Beit-Or says.

His advice: For deals, Beit-Or recommends hitting “the strip”—not the famous one on Music Row, but the area stretching from Chattanooga through Nashville to Clarksville. “These less competitive areas have better rent-to-price ratios and have quietly outperformed Downtown Nashville for years,” he says. “They also reap the benefits of Nashville’s strong economy and growth without the inflated prices.”

Beit-Or helped a client purchase a house in Clarksville listed for $325,000. But the seller wanted to stay put, so “My buyer ended up leasing the house back to the seller for one year,” he explains. In exchange, his client assumed the seller’s mortgage and snagged the house for $250,000. “Work will need to be done once the old owner/tenant moves out, but right now our client is looking at $75,000 in equity, a mortgage at an unheard-of rate in today’s times of 3.5%, and no vacancy with the tenant paying $100 per month over market value.”

Got a home or housing market you want to highlight in The Playbook? Tell us more about it here, and we’ll consider featuring it in an upcoming issue.