AI and 3D Printing Disrupt Dentistry: Align Technology Files IPO with $8B+ Valuation

This article is republished from Taishan Hui. The original title was “Many Doctors Will Lose Their Jobs to AI in the Future, and Dentistry Has Already Been Disrupted—Zheng Lixin of Demao Capital Discusses Medical AI.”

On the afternoon of July 25, 2017, the high-end salon titled “Medical AI: New Technology Disrupting Healthcare,” hosted by Taishan Hui and co-organized by Siyu Research Institute, Taizhihui, and VCBeat, was successfully held. More than one hundred investment elites and industry experts from the healthcare sector attended the event. Zheng Lixin, Founding Partner of Demo Capital, delivered a keynote speech titled “New Technologies Transforming the Healthcare Industry.” The following is a summary of his speech:

I am delighted to share my perspectives on artificial intelligence in healthcare with you. Today, I will begin by introducing Invisalign, a company in our investment portfolio.

There is a very famous dental aligner company in China called Invisalign., which is an artificial intelligence company based in Silicon Valley. One of the founders, Charlie, was originally a renowned algorithm specialist in Hollywood and participated in the production of the famous film Shrek. Many years ago, when Align Technology encountered computational challenges while developing its dental algorithms, it recruited Charlie to address the AI algorithms involved in tooth movement during orthodontic treatment.

Currently, over 90% of children worldwide have misaligned teeth. In China, only 1% of children with malocclusion receive treatment, whereas in Western countries, the treatment rate exceeds 90%. During orthodontic treatment, doctors bond individual brackets to the child’s teeth and connect them using a titanium alloy archwire. This procedure is known as orthodontics.

Orthodontics is the most technically demanding specialty in dentistry, as it requires manual dexterity to adjust titanium alloys. Orthodontists in the United States command very high incomes because it takes five to ten years to develop this tactile expertise, earning them the title of clinical experts within dentistry. Charlie realized that artificial intelligence could possess expert systems and expert logic, and that dentistry, particularly orthodontics, could leverage computer algorithms to replace these expert systems.

When Align Technology first proposed this concept two decades ago, it was met with widespread opposition across the dental community. All dental associations maintained that it was absolutely impossible for algorithms to replace the tactile expertise accumulated by dentists over ten years of practice. Yet, unexpectedly, Align Technology succeeded—Charlie made it happen.

This product is actually very simple. An algorithm can be used to design a custom aligner, and wearing this aligner for 24 months can correct tooth alignment. The aligners are manufactured using 3D printing technology. Over the past two decades, the company has grown to a market capitalization of $8 billion, relying on two core technologies: AI-driven dental modeling and 3D printing.

These two technologies have cracked the centuries-old techniques of orthodontic treatment and the authoritative expertise possessed by clinical specialists. As a result, there is no longer any need for clinical experts; merely a 3D printer, computer software, and a 3D scanner are sufficient to displace approximately 100,000 professional dentists worldwide.

If your child requires orthodontic treatment, it will take approximately 15 days to find a dentist.By leveraging artificial intelligence, the issue can be resolved in approximately five minutes.This is a case in point of the revolutionary changes brought about by the application of artificial intelligence in clinical medicine.

Future dentists are likely to be replaced; this has already been achieved in the United States.. Intraoral scanning, AI-driven design, and other tasks are completed either in factories or clinics, or even in supermarkets near your home; after 3D printing, the devices can be worn directly, with no involvement of dentists throughout the entire process.

So when many people meet with me, they say that artificial intelligence is a hot topic and ask whether it works. In fact, it is no longer a question of whether it works, because we have already placed countless dental aligners in the mouths of consumers worldwide.

I am using this example to illustrate that artificial intelligence is no longer confined to theoretical discussions but is already disrupting the healthcare industry. AI is not merely a topic of conversation; it has already seen extensive application in clinical practice.Invisalign has provided AI-powered orthodontic treatment to 4 million patients, generating $1 billion in revenue, achieving a net profit of $190 million, and reaching a market capitalization of $8 billion.

To what extent has this technology been applied? You and your spouse can go grocery shopping or stroll through the mall, and during the outing, your teeth will be scanned. The aligners will then be printed for you to take home and use, a process that completely eliminates the need for dentist involvement.

In 2006, Professor Yan at Tsinghua University, known as the "Father of 3D Printing in China," co-founded a company with President Wang Wenkang of Capital Medical University, calledAngelalign. In 2006, we invested in him. At that time, I didn’t even know what artificial intelligence was, let alone algorithms or 3D printing.

I visited Tsinghua University and spent a long time examining this technology, yet I still didn’t fully understand it. However, the two elderly professors with gray hair appeared highly trustworthy, so I believed it was worth investing in and proceeded to do so. Over the past century, orthodontic treatment has relied on dentists’ tactile expertise and the use of titanium alloy wires. This company eliminates that need: we employ 3D scanning, followed by 3D printing, to produce appliances that are fitted directly onto the teeth.

In 2015, this company was acquired by a CEO for RMB 600 million. Currently, it and Align Technology (Invisalign) together command a 90% market share in China, with Invisalign as the market leader and Angelalign as the runner-up. This case exemplifies the mutual intersection and disruption between artificial intelligence and the healthcare industry in China.

After investing in Angelalign, we identified another orthodontics company in Guangzhou. We invested in this company three years ago because it utilizes metal 3D printing rather than resin-based 3D printing.

Furthermore, as many of you are aware, Meitu has gone public in Hong Kong. Meitu transforms your appearance from genuinely unattractive to artificially enhanced; yet, upon meeting in person, the underlying reality remains unchanged. In contrast, numerous dental and clinical technologies can genuinely enhance the aesthetic appeal of a woman’s jawline, achieving a transformation from genuinely unattractive to genuinely beautiful. By leveraging algorithms combined with 3D printing technology, we can basically perform minimally invasive or non-invasive cosmetic procedures.

Previously, the largest dental manufacturing plant was located in Dongguan, employing 3,000 workers. With the advent of 3D printing technology, all factory workers have been replaced by machines. In the future, an artificial intelligence (AI) industry chain may emerge in dentistry. This ecosystem will integrate oral scanning data, large-scale CT scanner data, dental model data, and treatment records into a comprehensive data platform. By converting clinical techniques into logical frameworks and then into algorithms, procedures such as prosthodontics, orthodontics, implantology, and facial aesthetics can be completed simply by inputting data and executing straightforward 3D printing processes.

Dentists earn high incomes, but this technology will leave hundreds of thousands, if not millions, of dentists worldwide facing unemployment. Artificial intelligence is thus disrupting the dental industry.

We are currently negotiating a project with Meitu. While Meitu digitally enhances female beauty in virtual environments, we can deliver tangible, real-world results. For instance, an expert in Guangzhou charges RMB 200,000 to reshape the lower third of a woman’s face using orthodontic aligners. What does this imply? In the future, consumers will be able to upload their 3D images; through our design process, custom aligners will be delivered directly to their homes via courier, reshaping the lower face and correcting dental alignment. This transformation turns dentistry, a long-established clinical specialty, into a consumer-driven industry.

Moreover,Two weeks ago, Apple introduced 3D scanning technology, enabling consumers to directly replace many specialized tasks and techniques required in clinical settings.

Let me share a bit about my own background. I joined McKinsey & Company in Canada in 1994, where I served as a consultant specializing in the healthcare industry. As McKinsey was the global strategic advisor to Johnson & Johnson, my career path took me from Canada to the United States, and then to China. After leaving McKinsey, I worked at Sino-American SmithKline, Glaxo, and Jiangzhong Pharmaceutical. I spent many years in pharmaceutical sales before transitioning into investment. To date, we have cumulatively invested in more than 50 companies, focusing exclusively on the healthcare sector. Our portfolio spans active pharmaceutical ingredients (APIs), chemical drugs, large-scale medical devices, and small-scale medical devices. Among these 50-plus investments, approximately ten are in niche market segments.

Our investment philosophy is straightforward: focus on a single sector, establish deep roots, achieve comprehensive coverage, and cultivate a self-contained ecosystem. This year, we have also launched an industrial fund dedicated to medical AI.

Reflecting on a century of medical history, I would summarize it in one sentence: Over the past century, the development of healthcare has been a process of increasing specialization. Around 1900, a single physician and a nurse could provide care for all patients. However, as medical disciplines became more specialized, professional diagnostic and therapeutic methods emerged in rapid succession. By around 2010, an average patient visiting a hospital interacted with 19 different medical specialties.

Whether it involves biochemistry or medical devices, all these specialties are required to contribute to his diagnosis. This results in an excessively high allocation of medical resources and highly inefficient service delivery for a single patient. Ultimately, there is only one solution: leveraging artificial intelligence. We believe that cloud integration may occur before 2020, during which process many physicians will lose their jobs.

The current difficulties and high costs of medical treatment stem from many doctors’ “three unknowns.” First is “not knowing the drugs”: with 6,000 medications available, doctors cannot grasp the characteristics of each drug due to memory limitations. Second is “not knowing the diseases.” Third is “not knowing the patients”: a single doctor may treat 70 patients per day, and upon their arrival, the doctor sees only a few words in the medical record, leaving no opportunity to explore the patient’s background at all.

The lack of comprehensive knowledge has led to low precision in medical care, resulting in the waste of medical resources and overtreatment. Now, artificial intelligence can be used to calculate and tailor suitable medications, appropriate dosages, and the right doctors for each patient.。

Doctors are now overwhelmed. In 2016, medical information doubled every 3.5 years, meaning that by their fifth year of medical school, students found what they had learned in their first and second years already obsolete. By 2020, at the current rate, this doubling time had shrunk to just 73 days. What does this imply? No student can keep pace with the rapid evolution of technological frontiers through rote memorization alone during their training.

How can we build an intermediate super medical platform or super healthcare platform? A Chinese academician at the United Nations has been advocating a slogan: “Super Integration.” With the emergence of Industry 4.0 technologies, how can medical or healthcare information and technologies be integrated into a highly convenient solution for consumers? This development is likely to materialize within the next five to ten years, and such transformation is already underway in dentistry.

What personally fascinates me most is swarm intelligence. In the future, when a patient lies on the operating table, as many as 50 miniature robots may be deployed to diagnose and treat them—this is the envisioned convergence of future medicine.

Medical robots are categorized into large, medium, small, and micro sizes. The large-scale robot is the da Vinci Surgical System; medium-sized robots are designed for specific diseases or anatomical regions; and small-scale robots refer to microscopic nanorobots. In future surgeries, multiple robots may operate simultaneously within the human body, each assigned distinct tasks and working in coordination.

AI products are generally categorized by medical specialty, such as dentistry, cardiology, and fundus imaging. In the first half of this year, Demao Capital primarily focused on investing in AI-based medical products.

Jack Ma acquired a stake in Lianhua Supermarket two weeks ago, becoming its second-largest shareholder. Do you know what he intends to do? He is not selling products; he is selling services. Lianhua Supermarket stores can subsequently serve as touchpoints for Alibaba to deliver value-added services. For instance, the installation of 3D scanners in these stores would enable a wide range of applications. The maturation of business models, coupled with technological advancements, gives me sufficient confidence to believe that healthcare over the next five to ten years will be fundamentally different from what it has been in the past.

In the field of precision medicine, we do not make investments. The healthcare industry has two major development directions: one toward clinical medical care and the other toward industrialization. Without a team that includes two or three elite PhDs and possesses a solid foundation in genetic theory, investing in precision medicine is relatively challenging. Therefore, Demo Capital focuses exclusively on the industrial-scale direction.

In terms of sensors, you can envision that the entire healthcare industry will be dominated by them in the future. For example, the wearable devices we are currently investing in are essentially non-invasive (external) sensors. This field holds substantial future opportunities, and we have a dedicated team focusing on identifying investment targets within this sector.

I believe the development stages of medical devices are quite comparable to the five levels of autonomous driving. The first level is simple assistance, the second is task-specific automation, the third is conditional autonomy, the fourth is decision-making capability, and the fifth is full autonomy by AI robots. Currently, I think healthcare remains at levels one and two, not yet reaching levels three or four, but it is expected to continue advancing in the future. Therefore, I believe that future medical robots will encode top-tier medical logic into algorithms and execute surgeries through mechanical means, which is analogous to autonomous driving. Thus, my former employer, McKinsey, considers autonomous driving to be the primary application product of artificial intelligence, with healthcare ranking as the second.

Regarding the iPhone and Apple Watch, I believe the latter holds significant value. This morning, we just reviewed a pediatric myopia detection system that utilizes a wrist-worn device, which is quite impressive. To illustrate, three years ago in Switzerland, I encountered a "one-drop blood" project involving highly miniaturized chip technology. What does this entail? Consider a scenario where you are playing basketball and suddenly collapse. Under normal circumstances, you would be transported via ambulance to the emergency room, where blood would be drawn for biochemical tests to determine the underlying cause of the fall. However, the critical window for rescue typically does not exceed ten minutes. Currently, the most advanced global "one-drop blood" technology enables on-site blood sampling and immediate result delivery by physicians at the scene. Therefore, what I am most bullish on is not the iPhone, but the Apple Watch. If Apple continues to advance its technology, it could potentially integrate a comprehensive suite of analytical and biochemical testing capabilities directly into the Apple Watch, thereby centralizing all your medical information on the device. Ten years ago, in Seattle, I observed a project involving pendants designed for the elderly. Each pendant contained the wearer’s complete medical history, allowing physicians to immediately determine the appropriate treatment upon finding the individual collapsed. In essence, this represents the core potential of wearable technology, epitomized by the Apple Watch. This explains why Huawei is focusing on this area: it may serve as the true entry or exit point for the future of healthcare. This is a domain pursued by major industry players due to the extensive technological integration required.

Finally, let’s discuss human-machine connectivity. The emergence of human-machine connectivity may pave the way for human-to-human connectivity. Machine-to-machine connectivity is absolutely essential; the swarm intelligence mentioned earlier exemplifies such machine-to-machine linkage. These three components will be rapidly integrated in the near future, with the pace depending on how quickly adoption can be accelerated.

AI + Human Augmentation: By stimulating specific nerves, it can regulate your emotions and enhance your intelligence, physical strength, and more. A year ago, while we were in Silicon Valley, over 30 artificial intelligence projects were submitted to us.

I often joke with department directors at many hospitals, saying, “You’re done for.” As we all know, there are global technologies and devices for quantified self that allow individuals to track their own health metrics. In fact, the next stage after quantified self is algorithmic self, in which multiple intelligent agents are deployed around you to perform calculations based on your medical data. So what will doctors be able to do in the future? I estimate that the future role of physicians will be that of mentors: first, providing care and compassion, and second, offering guidance.

Core Scenarios Within the Next Five Years: First, patient- and consumer-centricity; second, all-inclusive and user-friendly healthcare institutions; third, distribution through chain networks, community-based facilities, and shopping mall-style locations; fourth, family-centered care, delivering nearly 50% of medical services directly to the home.

“During my many years in Canada, I visited a hospital only once. In China, however, I accompany family members to hospitals on a daily basis. This situation, characterized by both time-consuming processes and insufficient technical expertise, must be addressed. AI essentially codifies expert knowledge into algorithms; once algorithmized, it can be replicated at marginal cost—down to as little as one yuan. Why? This is inherent to the nature of software. Therefore, across the entire healthcare delivery chain, I believe the integration of AI and the medical industry has entered an acceleration phase.”

Regarding medical consortia, I have observed a significant shift in Chinese policy. Shenzhen’s Luohu District has established the Luohu Hospital Group, consolidating all medical resources within the entire district under this single entity. As a result, annual procurement spending on large-scale medical equipment, which stood at RMB 2–3 million three years ago, reportedly dropped to just over RMB 800,000 last year. Why? Because of the medical consortium model. Medical consortia may take various forms and be structured in different ways, but they are poised to profoundly impact resource allocation and business models. We are currently engaging with several such consortia to help them integrate AI solutions. In the future, this trend, much like the Two-Invoice System, will bring about substantial transformations across the entire healthcare industry.

From an investment perspective, the three largest AI-driven sectors are automotive, healthcare, and the film industry. The AI potential in the automotive sector is estimated at approximately $500 billion, while that in the healthcare sector stands at around $580 billion. You may be delighted to hear this, thinking we have tapped into a larger market. However, this market has a drawback: it is highly fragmented. It demands meticulous, specialized cultivation; although the industry is vast, it is excessively subdivided.

In the realm of swarm intelligence, you can imagine a world war ten years from now. If China and the United States were to engage in conflict, it would be a battle fought by drones—a combat scenario devoid of human personnel. This area is truly at the forefront of technological advancement, yet we currently struggle to identify suitable investment targets.

The human brain is the final frontier. It holds a vast array of unknowns and cutting-edge discoveries, making it well worth in-depth exploration. Last year, Chen Tianqiao sponsored a laboratory in the United States, drawing criticism from many Chinese medical experts who accused him of funding foreign institutions instead of supporting Chinese researchers. The rationale is actually quite simple: there are only a handful of top-tier laboratories worldwide. If one aims to serve humanity, investment must be directed toward the best available resources.

Another area I have invested in is sleep diagnosis and treatment. China has the highest prevalence of sleep disorders in the world. There is a common misconception among Chinese people: if a child plays all day and falls asleep immediately at night, snoring, the mother is often pleased, believing the child is sleeping well. In reality, snoring is a medical condition. Currently, more than 80% of sleep diagnostics and treatments in the United States can be conducted at home. In the future, the home will become an ideal setting for physical recovery and health restoration.

We have one item on our agenda that I would like to discuss with everyone. Many of the leading models and business models for AI have yet to emerge. I believe there is an opportunity for us to explore two key questions: First, when will the technology truly mature? Here, I am primarily referring to the regulatory environment and operational models. Second, how will clinical scenarios evolve over time? As I mentioned earlier regarding orthodontic technology, a company in the United States offered a service where patients could mail their data to them, and they would 3D-print the appliances and deliver them directly to the patients’ homes. However, while the technology proved viable, legal issues arose. Because dentists in the U.S. are represented by strong professional associations, a lawsuit was filed against the company, alleging that its practices were illegal. The case is currently ongoing.

Regarding unicorns in the healthcare industry, it is recommended that everyone take some time to study iCarbonX.. Currently, iCarbonX has captured digital data across eight clinical domains, including genomics; it has also acquired the largest protein tissue company and the largest microbiome data company in the United States, as well as Tianjin’s largest organic chemical analysis company. Furthermore, it has established a comprehensive database by cataloging every cosmetic product available worldwide.

iCarbonX acquired eight companies last year, covering all clinical needs of the human body. I remember having dinner with them a year ago and saying, “Let us invest in you.” He replied, “You can’t get in,” implying that we were not qualified.

An AI company CEO made a remark that I found quite intriguing. At the time, an expert challenged him, saying, “You people in the AI field have 99% poor-quality, unusable data; only 1% is usable,” implying that senior clinicians in hospitals would not collaborate with them. The CEO responded cleverly, “True, but physicians from the post-70s and post-80s generations are willing to work with us.”

AI companies are unlikely to take over doctors’ jobs for now; the consumer market is the true future.. If we can successfully collaborate with the United States this time, we can instantly gain access to hundreds of millions of consumer-end users to facilitate intelligent treatment.