Expand podcast: Artificial intelligence vs human intelligence in ortho

This transcript has been reviewed by AI and may contain inaccuracies.

Daniel Ramos:
Welcome to episode seven of the Expand Podcast. In this show, I interview Matthias Peper, owner of TP Solution and founder of the TPAO Congress. To date, Matthias has treatment planned over 125,000 orthodontic cases for aligners, so he is a true authority and brings a very pragmatic perspective to this topic of artificial versus human intelligence in orthodontic treatment planning. Enjoy!

Welcome everyone to the Expand Podcast. Today’s topic is kind of a continuation of my last episode with Matthias Peper, where we discussed a lot about collective intelligence—how a treatment planning service can help a dentist or orthodontist—and artificial intelligence. We’ll explore how AI can work together with human intelligence to make orthodontics even better.

I have here with me today, in this episode again, my friend and colleague, Matthias Peper, from Germany. He is, as you already know, and as I mentioned in the last episode, but I will repeat: he is the mastermind, with his team, behind more than 125,000—and counting—orthodontic cases planned with orthodontists and dentists from all over the world. It’s important to mention that he’s a part of several aligner societies in Europe and worldwide. He is in charge—though not alone, of course, because he believes in collective intelligence—of the TPAO (Treatment Planning Aligner Orthodontic) Congress, the best aligner congress in my opinion. This is my personal opinion, but I believe many dentists agree with me.

Matthias, welcome! Thank you so much for being here with us for another interesting and intriguing topic: how AI can help make orthodontics even better.

Matthias Peper:
Thank you so much for the invitation, and I’m really looking forward to continuing the discussion with you. Yeah, let’s start with it!

Daniel Ramos:
Thanks, thanks for being here. My first question, to get us started:

I like to watch movies, and I’m creating an analogy here. Orthodontic AI—is it a friend or the Terminator for orthodontists? What do you think? Do you think artificial intelligence can replace a human being when it comes to properly planning an orthodontic case?

Matthias Peper:
Thank you for the question. I definitely, personally, think it’s a friend. Why? Because the more artificial intelligence is used, the fewer administrative tasks I need to do. That’s what I personally believe. I can focus and concentrate on explaining to the artificial intelligence, or the system, where I want to go, and I save a lot of time. I’ll give you a very good example.

I started very early with digital treatment planning, even before I started my company, which was 10 years ago. So, 15, 16, 17 years ago, I was the manager of an orthodontic lab, and we started using software for very minor aligner cases. Back then, we needed so many hours and so much time to program the software, to explain what needed to be done—especially if we wanted to do movements that weren’t very compatible and had to be done sequentially. There’s software in the market now, like OnyxCeph and other software, which is beautiful. But if you look back in time, we needed many more hours to program and explain what we wanted to achieve.

So, I personally think AI helps us save time for the things that are important. And what is important? Making decisions and setting priorities.

Daniel Ramos:
Yeah, I completely agree. For now, and probably for many, many years, right? Even though this technology is advancing and growing exponentially, and we’re seeing faster, better solutions, we still need someone behind it to give the orders, right? At this point, we have several apps that can generate text, edit videos, and more. But different people behind them can generate different outcomes, depending on how they give the instructions.

Matthias Peper:
Exactly. It’s always about how we feed the software—what we input and how we explain things to the AI. That’s exactly what you said, and that’s how we get the outcome. But there’s one thing AI will never be able to do: make decisions like a human brain—make human, intelligent decisions. It’s heading in that direction, but I’m not afraid. I believe it will always be the human brain that makes the smart decisions.

Daniel Ramos:
I agree, although “never” is a strong word, especially when we’re talking about technology and how fast things are changing. I do think the way orthodontists practice will change—it already has—but I see AI in a positive light. I can see it saving time and helping with the hard tasks, like those that involve numbers, research, and data collection. It’s impossible for the human brain to be as fast as a computer in calculating force. Now, talking specifically about orthodontics—calculating the proper force for movement—it’s very complex. We don’t have a movement that’s just one thing, like rotation. It’s always a combination of movements, and calculating the exact force in each direction was impossible.

Matthias Peper:
Exactly.

Daniel Ramos:
So, what orthodontists usually did was use less force to avoid surpassing biological limits, and as a result, treatment took more time. But only with the AI we have now—using an aligner system with a software planning out the staging, calculating for each type of movement the proper amount of activation—can we be more efficient. We’re no longer calculating force manually. We have software to help us. So, more and more, the software will evolve to assist us. And until proven wrong, I agree with you. I don’t see AI as the “Terminator” for orthodontists, even though I love that movie. I think, for a long time, it will continue helping us to be better.

Matthias Peper:
And I would also like to add something. It’s both: the software is able to learn. In the future, AI will also be able to learn individually. For example, if your preferred protocol is semi-sequential distalization—50% with an aesthetic start and a frog staging anterior movement, just as an example—and you always request this specific protocol, after the fifth time you ask for it, AI might decide, “Okay, this is your protocol for future plans.”

This is one thing. The second thing is collective learning. Let’s say a doctor has a specific way of making a movement more predictable. Fifteen years ago, some doctors started asking for a specific protocol for retraction. This is a protocol that moves the anterior teeth with proclination first, then continues with intrusion, and finally with retraction. Now, the software can learn that this is a predictable protocol with fewer refinements, because it’s gathering a lot of data.

Matthias Peper:
Because we want to make the treatments more predictable, this is what artificial intelligence helps us with a lot. And on the other hand, it's exactly what you said—it really helps us save time because we don’t have to repeat everything again and again. The software, in my personal opinion, is not able to make individual decisions that only fit one patient. How should artificial intelligence know that this specific patient is not able to use elastics, or that this specific patient is not motivated to have an aligner treatment with more than 50 aligners? There are so many individual decisions, and even in the future, if we do not feed the software with these very specific individual situations, the software will not be able to calculate in the perfect way. So, in my opinion, it’s like yes, we feed the software with everything we can, but there are always individual parts because every patient is different and has individual needs. Also, every orthodontist may have a slightly different idea of how to treat the patient, so I’m really not afraid about it.

Daniel Ramos:
Hello, everyone. I’m just taking a quick break from this podcast interview to tell you about the DSD Aligners course. I host a live stream course where I teach dentists how to treat all their orthodontic cases with aligners, from simple to complex, with clinic-validated and science-based methods. Whether you are already using Invisalign or clear aligners, or you are a general dentist just starting out, this course curriculum covers the common challenges with proven systems. To get an idea of what the course is about, you can access a taster class free of charge on the webpage linked in the notes of this episode. Be sure to check it out, and I hope to see you at one of these upcoming courses. Now, back to the episode.

Daniel Ramos:
Taking what you just mentioned about the advantages for staging, for example, and many others, for those who are just starting with aligners and going to a congress to watch several systems, including in-office systems where they can buy the software and do everything themselves, do you think there is an advantage to using a company system because of the AI they are developing, compared to an in-office software today, for those who are starting now?

Matthias Peper:
Yes, of course. I believe that if you...

Sorry, let me clarify. Maybe you could repeat the question briefly so I can answer it better.

Daniel Ramos:
Considering that you mentioned company software—the software we are using to plan—is collecting our intelligence. They are using the protocols we are testing and validating, and they’re seeing which ones work better, right? And they are improving. So, company software mostly has this insight and artificial intelligence. Now, imagine a dentist or orthodontist goes to a congress and watches many speakers, each one defending their own system. There is a growing tendency toward in-office aligner systems, which means the orthodontist buys the software and plans everything themselves, maybe to save money or for other reasons. In your opinion, is there any in-office system that is collecting intelligence, or is this a disadvantage for those who opt to go this way? Is there a benefit to using the company system?

Matthias Peper:
Yes, it's exactly what you said. Especially if we talk about time-saving. We are completely independent, and we’re using a lot of different software. We use the software from big companies, and we also use some in-house software. There are different ones. The thing is, how much time do you want to spend on administrative tasks? Administrative tasks like bite setting, separating the teeth, then explaining everything—“Okay, first I want to do distal movement of 1.7 for three stages with 0.25 each, then continue with the six,” and so on. You have so many individual steps when programming everything by yourself that, for a complex case with in-office, in-house production software, my team and I sometimes need three and a half hours.

This is what I mean. If I use a software with many algorithms and AI included, we can reduce the time sitting in front of the computer. We reduce it to quality management and making decisions, and maybe just changing some strategies. For us, the treatment planning time is reduced by 90%, but it’s still necessary.

Daniel Ramos:
That connects to my next question. How do we take advantage of AI to help human intelligence be better? How do we use AI so that orthodontists can be better? That time you mentioned—three and a half hours sometimes to plan a complex case, inputting all these protocols into the software—can probably be reduced with a system using AI, a company system, to much less. Maybe to just one hour. The dentist or orthodontist who is listening to us needs to understand where they want to invest their time: in creating more human connections with their patients, investing time in better presenting the planning, explaining, treating, and monitoring. Creating this connection is what will differentiate us from a robot or a computer, right? Because if we want to fight the battle of planning better than a computer, that’s a battle we’re destined to lose. The computer is going to learn very fast, and if we want to do it ourselves, we need to be prepared for that challenge, right? It’s a lot of time and money invested in equipment and new software, so it’s tough.

Matthias Peper:
Exactly. And I would like to add something. In my opinion, it's about more than just algorithms and AI—it’s also about motivation. There’s a little difference between artificial intelligence and the algorithms of the software. The first and most important thing is that we have a huge amount of algorithms in the software that helps us save a lot of time because we don’t need to do all these administrative tasks, which are boring and not motivating. Then, there’s the artificial intelligence that learns from one orthodontist and improves over time, helping them save more and more time.

But in the end—and this is a very interesting part—it’s about setting priorities. What do I mean by that? It’s easy for a software or an AI to create a calculated ideal setup, but this setup isn’t about priorities. It’s just trying to bring all the teeth into a perfect end position. This is the biggest problem. What is the real priority? Why does your individual patient come to your office and want your support? Because they have very specific, individual needs.

Daniel Ramos:
Yes, well, time is important. I don’t know if you have one, but I have an iRobot—the Roomba, you know, the vacuum cleaner. The amount of time it saves us for a simple task is amazing. I can leave it running while I leave home, and I don’t need to spend an hour helping my wife clean everything. I can use that hour for something else.

Daniel Ramos:
Other things... okay, this is how I feel about artificial intelligence for dentistry, right? We need to use it to save our time, to spend that time better on things that really need us, right? Sometimes it can be for a noble reason—to spend more time with family, to take more time on vacations because we know things are going well, or to take some time to make the right decisions for the patient, to spend time on the real important decisions, and not lose time on something that's not important.

Matthias Peper:
Exactly, because AI or the algorithm can do it for me.

Daniel Ramos:
Exactly. And we know that, specifically for aligners, things like staging, attachments, and the shape of the plastic—more and more of these things are going to be calculated by artificial intelligence. No human brain can do this as fast. We need to use it, but besides...

Matthias Peper:
Exactly, yeah. And it's also a great innovation that’s going on at the moment. For example, 10 years ago, you really needed to make decisions like, “Okay, I’ll place an attachment here—hopefully it's not too close to the gingiva, or too close to the antagonist, to the upper premolar, for example.” And is the shape right? From most of the treatment planning we do now, we don’t need to think about these minor details as much. It really helps with quality management and quality control.

Daniel Ramos:
Yeah, and that’s exactly my next step here. We’re using artificial intelligence to plan and get the aligners delivered so we can start treatment, but it doesn’t stop there, right? Probably for the other tasks we have, it’s even more important and saves even more time during monitoring. You know a lot about this—we already met at an important congress in Paris to discuss digital and virtual care and monitoring. I left amazed by all the possibilities we have to monitor patients remotely, to prepare ourselves better for what's going on with the patient, even without them being physically present in the clinic.

What I’ve realized is that patients are coming in less, but they feel more cared for. They feel like we’re keeping an eye on all the changes with their aligners. It’s a very important moment in treatment, and I know your TPS is similar to DSD TPS in that you help dentists through the entire process. How do you feel artificial intelligence can help with monitoring? How can we use it to prevent the loss of time, avoid wasting time, and take advantage of it for other reasons?

Matthias Peper:
It’s really interesting what's going on at the moment. There are different ideas of how to do digital monitoring, and there are different applications and apps to use it. You can review step by step what happens during treatment. You can compare, for example, the refinements with where you started. We can even integrate things like CBCT scans, so you can particularly understand what happens during treatment. You can identify problems—like, "Oh, there’s an issue with the gum. Was it because of a wrong movement?" Then you can compare and see that no, there was no movement at all—maybe it’s the patient using their fingernail to remove the aligner. You can better identify where problems are coming from, and that’s the perfect way to ensure quality.

Daniel Ramos:
Yeah, one thing that impressed me was that we can personalize these monitoring software programs. Patients can use their own smartphones at home, and since everyone nowadays has a good smartphone, they can capture images or videos with their aligners on or off. The software then does all the work—it calculates if the teeth are moving properly. If not, it informs the patient to continue wearing the aligners. We can even personalize it for each case. If the patient isn’t using the aligners properly, they’ll receive a message to wear them more. If they lose an attachment, the clinic team gets notified to schedule an appointment. We are much better prepared to deal with patients this way. We reduce the waste of time, plastic, and resources—a lot of things we’ve already discussed in this podcast, including the environmental impact of plastic waste. By controlling the patient better, we can reduce treatment times and prevent unnecessary stages and refinements, which makes a big impact. We need to use these kinds of tools, right?

Matthias Peper:
Maybe it’s the first time in orthodontic history that we have a real system for quality management that’s 100% clear and step-by-step. Imagine having a system where we can truly manage quality. Picture this: you have three chairs in your clinic with three dentists or hygienists helping you. The software tells you that in chair number one, you’re losing more attachments than in chairs two and three. It tells you that in chair number three, the second quadrant is where attachments are falling off more. You can then investigate and solve the issue. Without software and AI collecting this information, it would be very hard to know. Using virtual monitoring systems, you can start understanding much better what’s going well, what’s not, and what needs to change. The impact at the end is incredible, right?

And we haven’t even touched on other artificial intelligence tools like ChatGPT and others, which we use to write better texts and communicate more effectively. I’m learning a lot from this. AI helps in meetings like this one by summarizing our discussions, creating topics and titles, and so on. More and more, we’re going to use AI to communicate with our patients, collect information, and protect ourselves. AI can be used for pretty much everything nowadays, right?

Matthias Peper:
Exactly. I’ll give you another example. We have our TP Academy platform, where we offer more than 50 hours of high-quality aligner education material. We start with tutorials about software and go very deep. Now, we can use AI translation tools to help us translate everything. For example, we’re currently translating everything into Spanish so that only Spanish-speaking doctors can access all this information. It’s amazing how easy it is to do this now, thanks to AI. It allows us to help even more markets.

Daniel Ramos:
Yes, and as we’re discussing here, we’re realizing that artificial intelligence is helping with diagnostics, planning, and treatment monitoring. But what about when the treatment finishes? How do you think AI can help us maintain the good outcomes we’ve achieved and keep patients aware? As a TPS provider working with many clinics, are you using any tools to help in the next phase, since retention is still an important part of treatment? What kind of technology do we have available?

Matthias Peper:
What we’re seeing is that more and more doctors are using virtual monitoring devices, like Dental Monitoring, for example. It’s very interesting because the patient, even after treatment, can easily use a special app...

Matthias Peper:
Mobile phone... to make photos and videos, and it’s amazing how precise the artificial intelligence can be in identifying if there are any changes, if things are going in the wrong direction, or if the patient needs to go back to the orthodontist. At the same time, the software will give an alert system to both the orthodontist and the patient if there are any problems. That makes everything much easier and provides much better quality control.

Daniel Ramos:
Comparing to what we used to do in the past, usually my protocol in my clinic, as soon as I finish an orthodontic treatment, is to ask my patient to come back sooner—within about one or two months—so I can check if everything is going well. Then, if everything is okay, I’ll schedule the next check-up for three months, then six months, and then once a year. In one year, a lot can happen, right? And even with the patient coming to my clinic every three months, sometimes the differences you and I can detect only show when the problem is already there, and then we have few options. We either need to do a refinement or we’ll have to accept that there’s been some relapse, right?

Matthias Peper:
Yeah, with this virtual monitoring system, the software can detect even the smallest changes or discrepancies. Then, we can use this information to, for example, ask the patient to wear the retainer more often. If everything’s going well, we can allow them to wear it only at night. But if we notice even minor differences, we can say, “Okay, it’s not quite ready yet—you need to wear it more.” And we have the data, we have the information there.

Daniel Ramos:
Yes, and on the other hand, what we’re also seeing is that especially the younger generation is becoming more flexible to work from anywhere because the world is going more and more digital. So even here, it’s much easier to maintain good control through monitoring. If your patient is abroad, living in Bali, or perhaps in Ibiza for a few months, how could you monitor them in the past? Sure, you could send them to another orthodontist for a check-up, but that’s not a quality management system. You might not need extreme accuracy, but you at least have a good idea of what’s going on.

Daniel Ramos:
And how much time do you think we save during a regular treatment with virtual monitoring, compared to a treatment without AI or virtual care? How much less do you think a patient needs to come to the clinic?

Matthias Peper:
Yes, I think with digital monitoring, it's about 50% less, personally. In some cases, it could be 70 or even 80% less. It also depends on how comfortable the doctor is with letting go of some control. But yes, you can do a lot of monitoring digitally.

Daniel Ramos:
Let’s go with 50% so we don’t exaggerate. Seventy percent is probably for a doctor who’s been using it for a long time and has a compliant patient. But at 50%, imagine opening up 50% more time—you’re doubling your time without adding any new equipment or chairs. You’re opening up time because your patients are coming in half as much as they used to. Just by doing that, you can decide if you want to spend more time in Mykonos, right? On vacation, maybe? And from Mykonos, you can monitor all the data from your patients via virtual care. Or, if you want to increase your business, you can open up more space for new patients. So yeah, this is another amazing advantage of artificial intelligence.

Matthias Peper:
Exactly.

Daniel Ramos:
We’re talking about good outcomes, quality control, diagnostics, planning, but in the end, the goal is the patient, and dentists need to run a successful business. For me, this is a no-brainer. This is something everyone should invest in. AI that can save your time, make treatments better, and make patients happier. No one likes to come to the clinic all the time. I used to think my patients loved visiting me, but I realized that was a mistake. They don’t want to leave their homes or take their kids out of school, drive an hour, and spend time traveling to appointments. With AI, I can spend 5 or 10 minutes reviewing their aligners remotely and say, “Okay, you’re good.”

Matthias Peper:
I agree. That’s an internal mindset that we have—we think things need to be a certain way, but that’s only our perception. It’s very interesting to understand that, especially with the younger generation. They want to do things the easiest way possible, and they’re used to it.

Daniel Ramos:
Exactly, and not only the easiest way, but with real quality management. With AI, we have facts. The system tells us if something is right or wrong. We have a great package here, Matthias, because for those who are afraid of technology or AI replacing them, what we’re showing is a package of benefits that can improve our treatments, save time, and let us use that time how we want—whether it’s playing tennis, traveling with family, or opening up more space for patients. It’s good for business. Patients love technology, especially the new generations. They already know everything about it, and we need to be prepared. This is a win-win for everyone. I see AI as our friend for a long time. Unless something drastic happens, but honestly, I don’t believe in that.

Matthias Peper:
Exactly.

Daniel Ramos:
So, to finish here, we’ve kind of already answered this, but I think it’s worth revisiting: What can’t AI do? Where do we still need human intelligence? What are the fundamental tasks where human intelligence is irreplaceable? What do you feel AI still needs to improve on? You don’t need to answer everything—I’m just opening up some ideas for your final thoughts on this very interesting topic. Where should AI developers invest more time to make it even better, so we can save more time and improve our experience even further?

Matthias Peper:
That’s a very good question, and it’s similar to the question we started with. I personally believe it’s also about our mindset. What does that mean? It means, am I afraid of change, or am I looking forward to it? Am I able to adapt? In life, it's so important to be open to adaptation and to see changes as something that can help us in the future. I personally believe that AI will always help us save time, by taking over administrative tasks that we’re not interested in. It will give us more time to use our brains for the right decisions, allowing us to dominate AI. Even looking at my personal life, today I’m very free—I have my laptop and can work from anywhere. All of my employees—my 20 technicians—can work from anywhere too. We were able to do home office with just our laptops, and we’re really using our brains together with orthodontics to find solutions for complex aligner treatments.

Even in 5, 10, or 15 years, I believe it will be the same. Maybe we won’t need one hour for one digital treatment plan—maybe we’ll only need 15 minutes. That’s how I see the future. I’m absolutely not afraid of it; I’m looking forward to seeing what kind of innovations will come.

Daniel Ramos:
Well, in my opinion, one of the things that everyone who knows DSD is familiar with is that AI cannot substitute us in a very important step during the process of diagnosing and planning: the second appointment. This is the moment when the dentist will present the plan to the patient, right? We have a journey, and we start with the emotional part—motivating the patient with the mockup.

We do this in almost every course, and I must confess, every time I see the patient seeing themselves with the mockup, they get very emotional, and I almost cry with them. This is a life-changing presentation for the patient sometimes. And this connection, the way we present and personalize the experience, which is a consequence of a lot of artificial intelligence beforehand, is something I don’t believe AI can substitute.

This human connection that we need to have during the treatment—the empathy—we need to understand what the patient needs. As you mentioned, each patient is unique, and we need to capture this uniqueness and personalize the experience. Then, of course, we use as much artificial intelligence as possible to reduce time and make the experience better, allowing us to invest time in the things that really matter.

And this is my final message: to not be afraid of artificial intelligence. It's not a Terminator, it's more like WALL-E—this cool, friendly robot that’s there for us, saving us time, a good friend.

Daniel Ramos:
Matthias, please, I’ll allow you to finish since this is the topic you brought to us. I really liked it, and I’d like you to give the final message to everyone listening. A final sentence before we finish this amazing episode with my good friend, Matthias. It’s all yours—the final message, my friend.

Matthias Peper:
Yes, thank you so much. I’d like to summarize it in a very similar way. Again, I personally believe that the most important part is our personal mindset. We need to be open, and we don’t need to be afraid. We just need to see the opportunities that allow us to do even more of what we love to do.

Daniel Ramos:
I completely agree, and I invite everyone listening to change your mindset and be open to these technologies. Thank you so much, everyone, for being here on the Expand podcast today with Matthias Peper. Thank you so much, Matthias.

Matthias Peper:
Thank you so much for the invitation.

Daniel Ramos:
Until the next episode, bye-bye.

Matthias Peper:
I’m looking forward to it. Thank you so much. Have a beautiful day. Bye-bye.

Narrator:
Thank you for listening. I hope you feel inspired to embrace change and see the rise of artificial intelligence as an opportunity to show leadership. In the next episode, I will be talking to Dr. Pedro Costa Monteiro about in-office aligners. Should the pilot build a car? Don’t miss it.