Somewhere about the sixth grade, I decided that I was going to ride the wave of the great sovereign profession of dentistry. I threw my surfboard in the water and surfed all the way to first base, 1987, as a doctor of dentistry. Then, around 1998, I began studying to earn an MBA.
In one of the first classes I took, I learned how most of the major moves of the stock market, from 1850 to present, were basically all technology moves. First, there was the telegraph. Then, with the invention of the steam engine, everything began to get faster, easier, higher quality, lower cost and smaller.
Whenever you had a very new, disruptive technology—like a gasoline-powered engine, measured in horsepower, replacing teams of actual horses—that’s where all the money went. As steam engines got smaller, they could use them to power ships. When the ships got to America, we started building canals. The telegraph turned into the telephone, which turned into the internet.
Early on, I realized that this AOL dial-up thing that everybody thought was a joke was actually a very serious game-changer. And because I already had my surfboard in the dental sovereign profession, I decided to add the next big thing—the internet—to what I was doing. I turned what had been a 4,000-subscriber, 30-page newsletter, The Farran Report, into an interactive community named Dentaltown, and then I was standing on second base. I’m still a practicing dentist who saw patients and did root canals and extractions this week, but I also have this dental internet media company.
Now I can see clearly who’s on third base—it’s artificial intelligence (AI), and it’s everywhere.
The big future for big data
When I bought my Cerec machine—before most of the people teaching CAD/CAM today were even born!—its use was limited because of the size of the processing chips. Back then, programmers had to decide between memory or speed of the microprocessor. Today, thanks to Moore’s perception (the number of transistors on a microchip doubles every two years, while the cost goes down), CAD/CAM and CBCT have become big, and we have a bunch of things all coming together now.
With elastic cloud computing, for example, a machine that normally would take a thousand hours to process something now can do it in an hour with a cloud computing service through Google,
Amazon or the like. Combine that with big data, which identifies algorithms, and dentistry’s really about to change.
When I scan an impression and send it to the lab, the scan works with just that one impression. But after 10 impressions have come in, the lab has 10 times more data. Move the decimal point over again, 100, and again, 1,000, then 10,000, 100,000, 1 million, 10 million, 100 million. The bigger the data, the better it figures out what’s going on. Imagine what a company that makes one of every 20 crowns in American could do with this much information.
Soon, a lab will be able to look at your scan and automatically rate how to proceed:
- For maybe the top 20% of scans in quality: “Yes, this impression is good: We’re going to trim the margins here, and it’ll go right to production.”
If the scan’s more in the 60–80% range, they might think, “Let’s send this back to see if Dr.?Howie wants to trim this margin himself.”
And if it’s under 60%, they’ll say, “We’re going to send this back to the doctor to say that it doesn’t look clear enough to trim this ourselves—would you care to reimpress?”
If it’s under 40%, they’ll say, “This scan stinks, and you need to rescan it, or you can send it somewhere else, or we’re going to charge you a higher fee because the odds we’ll have to do it again are high.”
Over time, if the lab ends up doing too many remakes for you, it might say, “Six out of every hundred crowns for you have to be remade, and we’re not playing that game. Now, instead of charging you $100 for a crown, we’re going to charge you $106.”
Additional uses for AI
AI is also being used to uncover insurance fraud: A company called Pearl checks images that have been submitted, and can identify whether an X-ray is a duplicate of something that’s already been submitted by other dentists in other countries.
Dashboard dental companies like Practice by Numbers can pull out your accounting and practice data and compare it to to that of its other clients to show you efficiences and problems.
Imagine how supply chain management can affect how you order dental supplies. With the internet of things going into big data, going into cloud computing, soon your supplier should be able to look at your practice calendar and say, “Dr. Smith is scheduled to take five impressions on Monday, and she doesn’t have enough impression material,” and it will automatically be shipped to you. No more humans coming by once a week, pulling tags and you running out of things.
Dentists might hate it if insurers get into AI and big data: What if they run your numbers and realize that 10% of your root canals are extracted in 12 months, so it’s not worth it to them for you to do their patients when an endodontist a mile down the road whose extraction rate at one year is only 1%? I like this transparency as a parent and grandparent—I don’t want my future generations suffering through bad dentists—but I expect a lot of dentists will have ulcers about this because they’ve never liked competition.
When all things become transparent, insurance companies are going to start telling you to really pick your specialties, because you’re going to have to demonstrate proficiency to qualify. That’s something you’ve heard me say many times: Don’t delude yourself into thinking you can do general dentistry and molar endo and implants and whatever else. Pick your specialties based on proficiency ... or AI just might.