Will AI Replace Dentists? A Reality Check with Prof. Falk Schwendicke

Dr. Khalil Karam (host) 00:00

We are going to host Dr. Falk Schwendicke today in our podcast series Next Dental.

Dr. Falk Schwendicke 00:04

People are getting more skeptical because they understand and see the power of the AI technology.

Dr. Khalil Karam (host) 00:10

How much of the data currently are validated?

Dr. Falk Schwendicke 00:13

The AI systems are currently oftentimes heavily biased. It might well be that in ten years there’s no need to be on site.

Dr. Khalil Karam (host) 00:19

Is AI going to take our place as dentists? I think we are already seeing it right now. Are we expecting self-evolving software?

Dr. Falk Schwendicke 00:26

Yes, because these chat-bots, as we know, hallucinate. I can post photos of what I ate for lunch today, but I would never post my electronic health record. We will need to embrace the technology and make it work for us.

Dr. Khalil Karam (host) 00:41

Hello, my name is Dr. Khalil Ibrahim and I’m from UAE.

Next Dental Presenter 00:45

Dr. Khalil Karam is a digital-transformation strategist. As creator of the MAS concept in Fujairah, UAE, he integrates technology with aesthetic care to transform patient experiences, shaping the future of digital dentistry across the UAE.

Dr. Khalil Karam (host) 00:58

We are going to host Dr. Falk today.

Next Dental Presenter 01:01

Dr. Falk Schwendicke is a professor at Ludwig-Maximilian-University of Munich and a global expert in AI and digital health in dentistry. He’s the author of over 500 scientific publications and advises leading organizations, including the WHO, FDI and ISO, on dental-AI diagnostics and health systems.

Dr. Khalil Karam (host) 01:19

…in our podcast series Next Dental. Dr. Falk, welcome to UAE.

Dr. Falk Schwendicke 01:24

Thanks for having me.

Dr. Khalil Karam (host) 01:25

Dr. Falk, as an AI expert, we are hearing a lot of things about AI, especially since the emergence of ChatGPT, and it’s becoming a culture of this era—like the music culture or food culture is trending to be AI.

Dr. Khalil Karam (host) 01:45

So everything’s about AI. What’s AI in dentistry and healthcare itself?

Dr. Falk Schwendicke 01:51

Well, first of all, it’s important to define what AI is. One definition is “what we do with machines that we formerly did with our brains”—recognition, seeing, talking, walking, and so on.

Dr. Falk Schwendicke 02:04

In dentistry, there are a wealth of applications. One is diagnostics—a big focus on image analysis—and we’ll discuss this later, but also speech and text analysis.

Dr. Falk Schwendicke 02:21

You mentioned ChatGPT as one driver. These large foundational models can analyze medical text or speech, converse with us or our patients, plan treatment, and—thinking of CAD/CAM—automate implant or prosthetic planning. In the end, AI will probably be present at every step of dental practice.

Dr. Khalil Karam (host) 03:01

So AI is an alternative, not a substitute. It will take many tools and functions—but, like a new team-member, we must understand it before working with it. Tell us about the beginnings of AI in dentistry and where we are today.

Dr. Falk Schwendicke 03:53

Basically, it started 20 years ago with new models from Google, Facebook, and other tech giants. The first medical impact was in dermatology and ophthalmology—skin and fundus images for melanoma or retinopathy—about seven years ago, and then things moved very fast.

Next Dental Presenter 04:22

MELAFINE (2017) became the first FDA-approved AI dermatology device, pioneering multi-spectral spectroscopy for melanoma detection.

Dr. Falk Schwendicke 04:34

Once that worked, everyone jumped on image analysis.

Next Dental Presenter 04:41

AI image analysis automatically interprets images, detecting patterns and features with high accuracy.

Dr. Falk Schwendicke 04:50

That’s what we did for six or seven years in all medical fields, including dentistry: “Give me an image and I’ll analyse it,” using machine learning trained on thousands of examples—say, a carious lesion on a radiograph.

Dr. Khalil Karam (host) 05:07

What boosted AI so fast in the last seven years? Computing power? Internet speed? We need to link those developments.

Dr. Falk Schwendicke 05:42

Three factors: data explosion (Internet-of-Things, social media—today we generate as much data in one day as from the dawn of humankind to 2008),

Dr. Falk Schwendicke 06:15

compute power (mainly GPUs—everyone talked about NVIDIA, briefly the world’s most valuable company),

Next Dental Presenter 06:25

NVIDIA’s market value exceeds $2 trillion (2025), powering the AI revolution with high-performance chips.

Dr. Falk Schwendicke 06:34

…and third, software advances: machine learning and deep learning.

Next Dental Presenter 06:47

Deep learning uses multi-layer neural networks to process complex data for sophisticated analysis.

Dr. Falk Schwendicke 06:56

These methods let software train itself—learning what’s in an image or a block of text.

Dr. Khalil Karam (host) 07:04

Are we expecting self-evolving software?

Dr. Falk Schwendicke 07:07

Yes.

Dr. Khalil Karam (host) 07:08

So we’ll get customizable systems that improve our ecosystems?

Dr. Falk Schwendicke 07:16

Absolutely. Dynamic AI self-trains; every user interaction makes it better. After a year in market, it’s excellent. In medicine it’s trickier—software is regulated; manufacturers must file metrics with the FDA, so day-to-day updates are harder—but in the long run we’ll see it.

Dr. Khalil Karam (host) 07:57

Regarding data volume and quality—how much is validated in AI for dentistry or healthcare?

Dr. Falk Schwendicke 08:20

Two issues: data origin and privacy, and what data we feed the AI. Dental data are tightly protected (forensic re-identification possible). We also know medical AI can be heavily biased—socially or epidemiologically—so we must verify it works on populations in Dubai or Germany, not just where it was trained (US, Korea, etc.).

Dr. Khalil Karam (host) 09:37

Philosophically: humans learn rapidly in adolescence—will AI show a similar spike?

Dr. Falk Schwendicke 10:01

We’re seeing it. Dentistry’s first AI apps appeared in 2020; now there are fifty or a hundred, getting better fast. The bottleneck is balanced data, but growth is exponential.

Dr. Khalil Karam (host) 10:36

Digital dentistry feeds AI; do we have enough adoption for proper machine-learning?

Dr. Falk Schwendicke 11:10

Dentistry is already quite digital. Most practices in high-income countries have electronic records and digital radiographs; CAD/CAM isn’t universal yet, but we’re generating enough diverse data to train AI.

Dr. Khalil Karam (host) 11:50

How reliable is AI now for automated tasks like segmentation or identifying novel structures?

Dr. Falk Schwendicke 12:16

For common cases—2-D or 3-D imaging, segmentation, caries or apical-lesion detection—AI works. Rare conditions (0.1 %) remain hard because of scarce training data. Remember, all medical AI is assistive by regulation; the final say is ours. We must also demand fair benchmarking datasets; that’s what we’re developing with the World Health Organization.

Dr. Khalil Karam (host) 14:13

Consumer trust in AI dropped last year; some fear job replacement. Will AI take dentists’ or technicians’ roles?

Dr. Falk Schwendicke 14:49

People are wary because tech advances so quickly—chat-bots, deep-fake videos, etc. AI will mainly be a helper. Communication-rich, diverse dental work won’t vanish. Auxiliary roles (e.g., lab-technician tasks already shifting with CAD/CAM) may automate further.

Dr. Khalil Karam (host) 16:01

We face a shortage of skilled technicians. Should new dentists memorise as much, or lean on AI?

Dr. Falk Schwendicke 17:00

We can’t ban tech—computers were once banned from classrooms! We must embrace AI but remain critical. Students using chat-bots must still verify content—these models hallucinate references. Otherwise we risk automation-bias (like motorists blindly following early satnavs into lakes). Users must know the limits.

Dr. Khalil Karam (host) 18:23

Ethically, social-media users share data freely; will healthcare data follow?

Dr. Falk Schwendicke 19:00

Probably not. People won’t publish their electronic health records. Healthcare data will stay regulated. Governments, though, realise affordable care needs AI, so they’re pushing for secure, shareable datasets.

Dr. Khalil Karam (host) 19:52

Cultural and biological differences vary by country—will AI need regional customisation?

Dr. Falk Schwendicke 20:11

Manufacturers dread that, but maybe yes. Dentists from different regions interpret the same X-ray differently. We may need AI customised to nations, systems, even individual users.

Dr. Khalil Karam (host) 21:18

Robotics: will we soon see AI-driven dentist robots in public?

Dr. Falk Schwendicke 21:59

Prototypes exist (a drilling robot backed by Mrs Zuckerberg’s father). Personally, I’d invest in an assistant robot—cheek retraction, suction, passing instruments—because many systems face nurse shortages. Robots will enter various roles, but dentists won’t be unemployed.

Dr. Khalil Karam (host) 23:03

Could one dentist operate multiple robots remotely?

Dr. Falk Schwendicke 23:17

Conceivable. With augmented- or virtual-reality interfaces, a dentist might oversee several robotic procedures remotely within ten years—scary but possible.

Dr. Khalil Karam (host) 23:50

Thank you, Dr. Falk. We’re out of time.

Dr. Khalil Karam (host) 24:03

Thank you, everyone—please subscribe and follow for more of these podcast series.

Next Dental Presenter 24:11

Want AI in your clinic? Try Remedico EMR, sponsor of this episode—pioneering AI for dental clinics. Visit Remedico.app for a free demo. Remedico: software for modern dental clinics powered by AI.

Next Dental Presenter 24:25

Hey—don’t miss the next episode! Search Next Dental (one word, no spaces) on all major podcast platforms and click “Subscribe.”