Friends of ResiShares -
As if downtown office landlords needed another reason to worry, robots are apparently coming to take millions of human jobs.
Robots don’t take days off, don’t get sick, don’t complain about their robot colleagues, generate images like the one above in seconds, and have even been known to do a servicable job writing this newsletter.
Most critically, they don’t need office space.
So the re-routing of VC cash from late-stage SAAS businesses to seed-stage AI businesses can be seen as unequivocally bad news for office landlords in traditional tech hubs. Right?
Perhaps not, at least according to JLL this past August, who saw demand for 700k - 800k square feet of downtown San Francisco office space from AI startups. Critics will point out that JLL is a commercial real estate broker, so is inherently talking its book, and that 700k-800k is barely a drop in the 25 MM square foot bucket of vacant San Francisco office space. Still their optimism does offer one of the few voices competing with the cacaphony of doomers predicting that AI will usher in a collapse of white collar employment opportunities.
And right in time for this month’s note, Sam Altman, the biggest voice in AI, corroborated JLL and signed the largest office lease in San Francisco since 2018. He got fired, then rehired, approximately 3 weeks later, and the AI conversation went back to a combination of Us Weekly office gossip, hagiographic descriptions of Microsoft’s Satya Nadella, and Millenarian doom prophesies of Open AI’s dysfunction portending the enslavement of humanity.
But seriously, though, AI is going to be a net creator of jobs, wealth, and office demand, and that SHOULD be the conversation we are having!
Back in 1996, two Stanford students wrote an algorithm that made it easy to type a few keywords into a text box and return relevant websites related to those keywords. Despite lacking any semblance of a revenue model, the pair was able to raise $25 MM in venture funding from Kleiner Perkins and Sequoia, the two most storied names in the sector. They called it Google (somehow less cheeky than Yahoo!).
In 1999, after that initial round of funding, and with Google’s flagship product launched to the public, the firm had a total of 13 employees, including its first executive assistant. Later that year, Marissa Mayer, the future CEO of Yahoo, would become their 20th employee.
Think of the tens of thousands of file clerks, research assistants, and other medium-skilled information workers around the country whose job was about to be replaced by a firm with 20 people and a powerful algorithm. Technology destroys jobs!
Then, in October of 2000, something happened at Google. They figured out their revenue model.
And with a revenue model came customers. And with customers came accountability. And a firm that is accountable to customers to make money suddenly realized that they needed to hire people whose job was to connect those customers to its products profitably.
And by the middle of 2000, Google had 300 employees.
And by the time they IPO’d in 2004, Google had 3,000 employees.
And then this happened:
Google now has over 180,000 employees. They earn something like 80% of their revenues from selling ads, most of which are driven by their core search algorithm, first conceived by 2 PhD students in their dorm room in 1996.
Who are these people? What do they do all day? Why do they need 180,000 people to mostly oversee an algorithm that already exists, and is largely monetized through a self-service platform (Adwords)? While it’s difficult to find out exactly how Google organizes its employees, I was able to get a bit of a hint with a keyword search of their employees on LinkedIn.
Unsurprisingly, of the 142k current Googlers on LinkedIn, the lion’s share of employees carried the word “Engineer” in their title, at 59,000. After engineers, employee titles looked…
…kinda like other companies look. While a rather unscientific way of going about this research, (think of all the “Product Designers” “Marketing Support Analysts” and “Sales Engineers”) just this brief study points to approximately 40,000 people at Google who don’t ever touch code.
Talking to customers and prospects, gathering intelligence about what they want, feeding that back into front end products and delivery systems that connect them to the company’s core technologies are all labor-intensive jobs, as are the HR, legal, facilities management, and cafeteria baristas that support them.
770 people currently work at Open AI, according to Wikipedia. If it really wants to take over the world, it is going to need hundreds of thousands of workers, inside the company, at 3rd party dev shops, and inside their customers, to render the power of AGI truly useful.
If your bank accidentally takes $2000 out of your account when you perform a $200 ATM withdrawal, claiming, “I don’t know how it happened, we outsource that function to Chat GPT,” is not going to be an acceptable answer. Building, maintaining, selling, supporting, and documenting the right chain of accountability for that process and millions of others, will be the job of hundreds of thousands of people.
And many of their rear ends will occupy seats in offices from 3-5 days per week.
Speaking of technology coming for jobs, we may finally be on the cusp of the big bang for real estate brokerage. The National Association of Realtors (NAR) just lost a $1.8 BN lawsuit for conspiring to keep commissions high. Inman has more details, including the allegation by electronic discount brokerage REX that a buyer’s agent refused to show her client a REX-listed home because they were not offering a guaranteed buyer agent commission. They actually caught her on tape saying that she will not show the house to her buyer, with whom she signed a contract, likely with the word “fiduciary” in it (or making reference to state regulations with that word in it).
Sounds pretty bad, no?
At the same time, I sympathize with this agent. Like, how was she supposed to get paid, other than by the seller, given the historical norms of the real estate market? Was she supposed to go back to her client and beg for money out of their pocket after the deal closed? It seems blatantly obvious that current technology should be able to facilitate a more transparent real estate market that allows brokers to compete on costs, following the evolution of every other large financial and commodity market in the world, but until that happens, no individual market participant has the ability to do something about it. Don’t hate the player, hate the game.
What does this mean for the brokerage industry going forward? It seems like the path towards a cheaper and more transparent market can be initially embarked upon either by the private sector or the public. For instance, a copycat of REX (now defunct) could emerge with sufficient financial backing to stand up a true two-sided marketplace in a high-volume geography, then scale its business model nationally. Unfortunately, given both the dearth of transactions and VC dollars flowing into proptech right now, this seems unlikely, but it is certainly possible that a player like Redfin, Compass, or Zillow could make this happen.
Alternatively, the industry, led by NAR, could decide to cut its losses and propose its own solution, regulated by a Self-Regulating Organization (SRO), akin to the way FINRA polices the securities industry. The reason that they would consider this is that they see the writing on the wall and want to at least take control of the transition to a potentially less lucrative paradigm instead of leaving it to the courts and state legislatures.
Either way, it’s always easier to make these changes in bad times than good, when there is less lucre over which to fight.
T-RECS Of The Month: Boston-Cambridge-Newton, MA-NH
Boston is a big, wealthy, and expensive city in a cold part of the country - a series of factors that has done no favors to its home owners over the past several years. By contrast, it offers the employment opportunities and world-class educational resources of peers like New York and San Francisco at a substantial discount, and with both indoor cultural and outdoor recreational opportunities in rare combination.
T-RECS thinks Boston is primed for significant outperformance in 2024, driving it to the 81st percentile of the top 200 US metros over the next 3 years.