The Biggest AI Productivity Gain Is Not Speed. It Is Knowing What to Do Next.

The biggest productivity gain from AI isn’t doing the same work faster. It’s knowing which work deserves your attention next. That’s what I’ve learned from building AI into the way I run my real estate business, and I think it applies to almost every workplace.

Most of the conversation about AI at work is still about speed. Can it write an email faster? Summarize a meeting? Build a presentation or a marketing campaign in minutes instead of hours? Those things are useful. But speed isn’t the hardest problem AI can solve. The harder problem is deciding what deserves your attention in the first place.

A buyer who had been looking in the wrong place

That became clear to me through a buyer who had already spent months searching for a home with another agent. They knew what they wanted, but nothing on the market was working.

We approached the problem differently. Instead of waiting for the right property to appear in the MLS, we used our technology to analyze properties that matched what they were looking for, even if those homes weren’t for sale.

Within a few weeks, we found the home they’d been trying to find for months. It was off market. They bought it, and the appraisal came in roughly $65,000 higher than the price they paid.

The most interesting part wasn’t that technology made our search faster. It changed where we looked. That distinction matters.

Information isn’t the problem

In almost every industry, people now have access to more information than they can reasonably process. Salespeople have enormous CRMs. Recruiters have thousands of résumés. Executives have dashboards full of metrics. Customer-success teams have years of engagement data.

The problem is rarely a lack of information. It’s knowing which piece of information matters right now.

A seller who needed the right buyers, not more of them

I saw the same principle work from the other direction with a seller. Agents traditionally launch a home by putting it in front of as many people as possible. We wanted to be more deliberate.

We used AI to help identify the kinds of buyers most likely to connect with the property, then built the marketing around more than bedrooms, bathrooms and square footage. We focused on the story of the home and the life it offered, and put that story in front of the buyers most likely to care about it.

The response was 18 offers. The home sold for almost $300,000 above the listing price.

AI didn’t negotiate those offers. It didn’t walk buyers through the home or understand what the property meant to the seller. What it did was help us decide where to focus.

What the research says

The best early evidence points the same way. In a study of 5,179 customer support agents, access to a generative AI assistant increased the number of issues resolved per hour by 14% on average, and by 34% for newer and less experienced workers, with minimal effect on the most experienced (National Bureau of Economic Research). The biggest gains came from helping people know what to do next, not from typing faster.

That may be the more important workplace application. A salesperson doesn’t necessarily need software to write another follow-up email; they need to know which five prospects out of 500 deserve a call today. A recruiter needs to know which candidates deserve a conversation now. A CEO needs to know which number on the dashboard requires action.

Finding opportunities before they reach the market

We use the same idea with real estate investors. Investment opportunities often begin long before a property reaches the open market. Public records can reveal situations such as probate, foreclosure activity or tax issues, where an owner may be considering a sale. Foreclosure filings are rising: ATTOM counted 227,548 U.S. properties with filings in the first half of 2026, up 21% from a year earlier, including 21,543 in California (ATTOM).

Finding those situations, and working out which ones are actually relevant, used to take an enormous amount of manual research. Technology can now narrow thousands of possibilities to a much smaller group. From there, the work becomes human again: respectful outreach, understanding what the owner actually needs, deciding whether a transaction makes sense, and possibly connecting that owner with an investor looking for that kind of property.

That human part carries real responsibility. California law exists precisely because homeowners in foreclosure have been “subjected to fraud, deception, and unfair dealing by home equity purchasers,” and it sets strict rules for those transactions (Civil Code 1695). Bank-owned sales have their own rules too: larger California lenders must give owner-occupants, affordable-housing nonprofits and public entities the first 30 days to make offers on foreclosed one-to-four-unit homes (Civil Code 2924p). For the investor, a well-handled opportunity can mean finding a property before it faces broad competition. For the owner, it can mean another option. Our investor page and our article on using AI to uncover off-market opportunities explain how we approach it.

Activity isn’t productivity

Companies still measure a surprising amount of work by activity: calls made, emails sent, meetings held, tasks completed. But 100 poorly targeted calls aren’t necessarily worth more than 10 well-timed ones. More activity and more productivity aren’t the same thing, and AI gives us a chance to rethink that relationship.

The best systems won’t simply help people produce more. They’ll help people decide where their time is most valuable. That doesn’t make human judgment less important. In my experience, it makes it more important.

Technology can identify a pattern, rank an opportunity or flag that a property or relationship deserves attention. It can’t fully understand why someone is selling a family home after decades. It can’t read the hesitation in a negotiation. It can’t build trust with a buyer about to make one of the largest financial decisions of their life. AI is very good at scale and pattern recognition. People are still responsible for judgment.

The better question

The companies that benefit most from AI may be the ones that stop asking, “How can we use this to make people do more?” A better question is, “How can we help people spend more time on the work that matters?”

The first wave of generative AI showed how quickly machines can create information. The next wave may solve the opposite problem: helping us decide which opportunity, customer, problem or decision deserves our attention. In a workplace already drowning in information, that may turn out to be far more valuable than doing everything faster.

You can see the same forces in the housing market itself. Redfin reported that San Francisco’s luxury home sales rose 22% year over year in March 2026, with AI-industry pay cited as a major driver (Redfin), and Compass found $10 million-plus sales reaccelerating nationally in 2025, with San Francisco and Silicon Valley among the markets reversing declines (RISMedia). I wrote about what that means for wine country in AI founders and Napa-Sonoma luxury real estate.

More about me and my work is on my profile and our media page, and you can read what clients say on our testimonials page. If you’re buying, selling or investing and want a different approach, get in touch.

Sources & notes

• Brynjolfsson, Li and Raymond, “Generative AI at Work” (NBER)
• ATTOM Mid-Year 2026 Foreclosure Market Report; Civil Code 1695; Civil Code 2924p
• Redfin: San Francisco luxury sales (May 2026); RISMedia: Compass ultra-luxury report

The two client results described are from Diamond Real Estate Group transactions; individual results vary. Last reviewed October 2026 by Matthew Martinez.