Transcription
When we don't fully understand a new technology, often times that can lead to an immediate distrust and really knowing how to use it.
Just seen a survey and it the the survey was from the Chronicle of Higher Education and they pulled faculty members in higher education. They said, you know, how comfortable are you with AI and 72% of them said not comfortable at all.
Statistic was 98.75% you cannot tell the difference. So whether the answer in the poll was given by you know artificial intelligence or whether it was actually somebody living sitting in New Jersey or Missouri whatever uh that was impossible to discern. That's that's the kind of thing to me that you just have to be aware of it. How do you put guard rails around something like that?
So as everybody is racing forward, I I totally understand the exuberance of just being a little bit older and having seen things I think the need uh and again both you and Will Par have talked about it in terms of not just the regulatory but how do you how do you promote the bioeththics or just the ethical forget about medicine and dentistry just the ethical behavior of you know, the whole landscape feels like this is the ultimate generational kind of moment if you will because we're at the precipice of something really big that is going to And according to most experts, this is going to be even bigger than the web.
Don't fall asleep at the wheel. Meaning, don't get left behind at the same under the same token, right? Don't let other industries adopt AI and dentistry gets left behind. And third of all, let's continue to be critical and and sort of from a regulatory and ethical perspective, just like pilots, you know, aren't absolved from incidents under an autopilot scenario, we cannot be absolved from any incidents when AI is being used in our clinical setting.
So to who are coming from across the world was incredible. You know, to have a poster next to me from Brazil, someone on the other side from Germany was fantastic to see, you know, where are they at? How are they thinking about things and coming into the space to do it.
...All right, let's go... Beautiful... Welcome to the Pulp podcast where we talk about all things dentistry and non- dentistry from a generational perspective. I'm joined today with my co-hosts Dr. Anne Koch is joining us from the beautiful Palm Beach Gardens Gardens down in Florida and Dr. Parsa Shahidi is here in the studio and we also have a new uh guest co-host here, Dr. Roene Nasr, I guess soon to be a doctor, but we consider all our, you know, dental students already doctors.
I'm honored. We're working hard out here. That's terrific. That's wonderful. And we have for the first time Parsa a repeat guest. And who could be a better guest than the guest that we have today. So why don't you go ahead and do a quick.
Yeah. I mean people know uh our our repeat guest today, but just in case you don't, which I doubt that's the case, we have our uh esteemed and beloved dean of Harvard School of Dental Medicine, Dean Gobi. Welcome back. Good to see you. How are you? Good to see you, Parsa. Oh, glad to be invited back. This doesn't always happen when I meet with people. So, thank you so much. We're happy to have you here today. And I think Dr. Nasseh, today we want to talk a little bit about what an amazing event we had just a short while ago at the second inaugural Harvard Dental AI symposium. Today's the day to talk about that and give our viewers a little bit of an idea what happened there. How was it?
Well, yeah, it was fantastic. It was really the energy that we had in the room and then all of the workshops, the research poster sessions, seeing our students, our staff, our faculty and people from 46 different countries. Wow. All over the world uh present with us in Boston. It was an exhilarating experience. Wow. 46. Uh it was incredibly well attended and I thought there was a great lineup of speakers, keynotes, and other people. Um Annie did not get a chance to attend. We were all the three of us were there. So I have a lot of questions. They're kind of macro questions. Um because I think also uh probably a little bit of a difference to the three people with you will I'm I'm still a little cautious about AI like a lot of people and then I'm going to talk about some of the things I have that's especially on some of the uh the valuations and things like that and how much money is going into these data centers and that kind of stuff but that's different than dentistry. So my real question to you about this is what when you put together this symposium, how did you make it? Did you make it totally macro general AI and dentistry or did you go into really specific dental things like you know diagnosis and endodonics or diagnosis and periodonics? How did you come to that decision how to make the format what it was?
Yeah, Annie, good question. And as we think about so this was our second global symposium on AI and dentistry and so as a part of this process we learned from our first symposium and we recognize there's a tremendous amount of energy excitement about AI and really it's it's across all industries and so what we did in our first symposium is really blend together leading keynotes from individuals who are in crosscutting areas and so we had this again with this symposium with Kareem Lakhani from Harvard Business School, Jim Weinstein from Microsoft Health is those two keynotes and then we had we brought on board students. We had about 90 different student presentations. Roene, she had uh one of them that was recognized and then we had uh workshops, hands-on workshops on using AI. And then the three focus areas to your question were on clinical practice AI uh which we know there's so much interest in this particular area and then looking at dental education in AI and then the regulatory landscape and the bioeththics and then at the very end we had lightning talks with people who had cutting edge uh information that they presented seven eight minute talks uh for areas of interest. So we felt like you know there was something for everyone. Um and at the same time as you said you know really trying to understand uh where we where we sit with AI right now because it's changing by the day and you know one of the things you know everybody's so siloed so soon as you have any disagreement with something right away you labeled a dinosaur or something like that or like then people will say like oh my god it's just like TV or it's like your regular computer. We all know it's certainly not anything like the cathode ray TV or your computer. But one of the things that you mentioned on my concern is not about what it can do in all these different fields. My concern is you you talked about it at the end of your your little intro there, Will, is the regulatory and the bioeththics. my my my biggest concern with AI and I think that's why having such a symposium as such a reputable institution is so important because to me as someone who is nowhere near as high-tech as the people around you I'm looking at I'm thinking I think there's an opportunity here for some unsavory behavior and and again maybe that's because I'm like totally ignorant in this but I think the regulatory aspect and the bioethical is something that really needs to be addressed Right.
Well, you know, it's something it, you know, when we don't fully understand a new technology, often times that can lead to an immediate distrust and really knowing how to use it. You know, I was just I I had just seen a survey and it the the survey was from the Chronicle of Higher Education and they pulled faculty members in higher education. They said, you know, how comfortable are you with AI? And 72% of them said not comfortable at all or just have a limited level of comfort in teaching it. And I'm to I'm totally there. Everyone knows it's important and we need to learn much more about it. And so when I see someone like Rowen, our younger generation, they're growing up with it and they're understanding it actually much better than the senior set of individuals being able to adjust to it. And so we could talk at length about the bioethics and regulatory landscape. uh one of our speakers at the symposium even said you know when we first had uh calculators and using them for trigonometry the everyone was worried that we would never learn how to do math and it's it's the same thing I think with AI we have to use it because I mean the way that they I think people are looking at it and Dr. got your point. It's are we, you know, we used to ask calculators to calculate things for us. Now we're asking AI to think with us. And I think what the symposium did and I think we touched on this a little bit is it set up like kind of the beginning of the symposium for those who didn't attend. The idea was that um we heard an amazing talk from Dr. Khani who had gave us a sort of overview of AI in terms of what's happening outside the field of dentistry across healthcare maybe a little bit of dentistry. It sort of opened up people's eyes to what's happening and that the theme of don't fall asleep at the wheel which I think we should talk about and then we got into these different sort of subcategories that I think you know regulatory and how to actually implement it computer vision diagnostics the research that we saw um what would you say you know to to all of you what were some of the highlights that you would tell tell our audience from from your experience at the symposium what was the best
well, you know, maybe I'll I'll I'll flip this to Rowen in terms of from the student perspective as a participant and a presenter. Totally. What were some of the aspects that really attracted your attention? I'd like to because everyone uh Annie didn't have a chance to attend, but just to hear your perspectives attending the meeting.
Yeah. No, I totally agree. I mean, to be in a space where so much is being talked about where we see a lot of it on the media, we get bits and snippets of it in class. I think our professors are starting to address it and recognize it because I think what our our school does such a good job of is kind of staying with what's going on and I can tell even professors are starting to implement things into the syllabus. It's being talked about in class absolutely but to be in a space where research was coming up to hear about the innovative kind of concepts that people are bringing in from all spaces diagnostics imaging um treatment all these kinds of things great to have a hands-on kind of place to engage with that as a student because I think I mean like you were saying I think what it looks a little bit different for us education even like coming into the clinic coming into the space I'm really curious to see how things will shift as our generation navigates, you know, entering uh like the the really the hands-on component of dentistry and not just like the classroom setting. So, being able to do some research, present it, get some feedback from people who are coming from across the world was incredible. You know, to have a poster next to me from Brazil, someone on the other side from Germany was fantastic to see, you know, where are they at? How are they thinking about things and coming into the space to do it. I think I actually was curious to to think about your guys's thoughts on on the flip end. You know, as students, I think instead of maybe a little bit of concern, I think we're just very excited and I think we should be, you know, there is some healthy level of concern that should exist. Um, but I think a lot of different generations were excited. But how do you guys feel about that excitement and that balance that we should and can strike with AI as it rises and kind of enters our spaces?
I mean, we're all working in different areas. I mean to me it feels like this is the ultimate generational kind of moment if you will because we're at the precipice of something really big that is going to come and according to most experts this is going to be even bigger than the web right so it's it's a really really big deal so Andy's concerns are very valid and that probably highlights why there is this generational gap which is the youthful optimism and and our generation obviously we've seen more things go wrong and so we tend to be a little more jaded but but What's interesting is that um I feel like this particular conference kind of reminded me of every time there is an iPhone release with the Apple, they always say this is the best we've ever done and but you know the next one's going to be the best one, right? Because it's just the kind of thing that is going to improve with time. We're at such a next level of this technology. There's no other way. It just it's going to get better and better every year. But I think for this particular one for me it was just a good sense of understanding that so many people as you said Rowen are interested coming from all over the world yeah to attend this conference it's gonna a moment of bringing everyone together to talk about something that is going to be uh a defining moment for all of us and I think while I could probably confess that you know there is still not too many things in dentistry that are related specifically to AI yet to the level that we would be expecting when you go to a regular conference that is But just the foundational work and creating the networks and all of that stuff is just so important.
What what about yourself, Parsa?
It's a good point because I mean I think it's so early and it's evolving so much that we know what we're we're getting will be different in a year. But I think what's so important is if you are an adopter of technology or digital innovation that you're there and you're experiencing sort of the giants of our industry or the giants across the AI field like Kareem Lani help us understand how to navigate it right and so one of the things that I one of the most important takeaways was this concept of don't fall asleep at the wheel right he talked about you know how u in in our space maybe we might look at it in four ways one is you know we cannot rely entire entirely on the things and the insights that AI is giving us. That's number one. Number two is don't fall asleep at the wheel. Meaning, don't get left behind at the same under the same token, right? Don't let other industries adopt AI and dentistry gets left behind. And third of all, let's continue to be critical and and sort of from a regulatory and ethical perspective, just like pilots, you know, aren't absolved from incidents under an autopilot scenario, we cannot be absolved from any incidents when AI is being used in our clinical setting. So to me being there and understanding how people are looking at it you know he talked about partnering with Proctor and Gamble and BCG and these massive institutions the financial services and engineering we can just learn how to adopt it better in our space I think and that was the that was a takeaway for me and I I'll just add this one other thing and and ask you guys how you feel but you know the international ecosystem of AI is especially across some of the Asian countries are moving they're moving so so so so asked and it's always interesting to talk with some of those folks and we had a lot of those kind of pioneers from South Korea and Japan and trying to visit us as attendees. Hearing how they're implementing the research in dentistry with AI is absolutely phenomenal. I don't know if you guys saw anything that was interesting to you about that or or if you've heard anything about that during the conference.
Well, one of one of the things that was a little bit concerning to me, I was reading an article last week and I was talking about research and they were talking about how polling can be so adversely affected. You know, when you have those are you a robot type of things? No, I'm not a robot. Uh that's easy to discern. But with AI and a real human, um 90 what was the stat statistic was 98.75% you cannot tell the difference. So whether the answer in the poll was given by you know artificial intelligence or whether it was actually somebody living sitting in New Jersey or Missouri whatever uh that was impossible to discern that's that's the kind of thing to me that you just have to be aware of it. How do you put guard rails around something like that? So as everybody is racing forward I I totally understand the exuberance u just being a little bit older and having seen things I think the need uh and again both you and will para have talked about it in terms of not just the regulatory but how do you how do you promote the bioethics or just the ethical forget about medicine and dentistry just the ethical behavior of you know the whole whole landscape right
yeah, I mean, it's a challenging area the last session of the symposium really focused on regulatory and bioethics and the food and drug administration is putting some of the guard rails in place and looking at how AI is used in particular for computer vision. Uh there are a myriad of companies that have come together on computer vision looking at radiographs. We understand there are more uh dental radiographs than there are medical radiographs and so there's so much digital information that's out there and really setting up those standards as it relates to uh that particular area but I think so many other aspects to it when we talk about uh dental education educational programs there can be this cognitive deskilling so this being asleep at the wheel over relying on AI to make some of these decision making uh aspects and so we have to be very careful with that. uh as you mentioned Annie uh now they have these uh call centers where it's AI bots they sound as you said very human right and if you look at from college uh admissions to now dental office staff and one of the compelling areas at the meeting was several people said within the next two to three years the back office of the dental office and even some of the front office aspect effects will be replaced by AI and and it's you know I think both you and I realize uh dental schools and and people like yourself deans of medical schools and dental schools have to set a standard for the bioethics. I'm not concerned about any medical and dental school professional school I'm concerned about the people who are on the other side of the equation. So when I when I was reading this about how polling is at jeopardy, that's a big issue, right? It's not just whether you like this movie or not, but when you think about how much polling is done in all different manners of research. So I I think as this things and again the process you said there trillions of dollars being put into things. I think the whole concept and the awareness of having some guide rails I think is paramount.
Yeah. I think speaking of big takeaways, one of the most interesting kind of statements made during the symposium was how as humans we are so used to thinking linearly and AI is completely breaking the way that we're thinking about things now cuz it is just growing exponentially and almost for us to conceptualize what things are going to look look like in the future is almost beyond us still even at this point to even consider what could 2 3 5 years look like. I mean, did you feel a difference between the first and second sympo symposium even in what you're seeing and do you have any thoughts on even what the next one might bring?
That's a great question. Yeah, I mean I it's a good question because just in two years right there was dramatic leap in terms of what is is out there. I do think that now with the consolidation of dental practices and DSOs. Oh yeah, these practices are worked now and many of them have AI already implemented. So seeing how widespread it is within the dental practice community is actually quite widespread and uh so that was a big jump and then just seeing the realities of so Dr. Lakhani showing the avatars of himself. You know, I may be replaced as a faculty member, but you know, if I I want to talk about a different topic. It was it was quite compelling to see those avatars. That was really nice. Yeah. But, uh, yeah, it's exciting, you know, and I I'm even thinking when you you ask about the change from two years ago to now. Yeah. and we're already going to be planning another one in two years, but we're going to start seeing this changing so much and AI will be involved in everything, right? You know, to have a standalone symposium on AI will probably not happen because it's so integrated into everything that we're doing. Every industry needs to have it. But at least right now, you know, we feel that we're in a quite a fortunate position at Harvard University. You know the Kempner Institute for Natural and Artificial Intelligence is really a leading AI entity on aspects relating to the natural intelligence neurosciences to AI and so much of it is in the support and training of the next generation. So with that, we still are in a wonderful uh climate and uh ecosystem here at Harvard that we can continue to hopefully convene right these programs in the future.
Yeah, that that that's been exciting. Actually listening to Bernardo in the previous meeting talk about the natural intelligence and how it comes together was exciting. But to your point Roin with the you know things are changing. Yeah. And then there is an acceleration of change and then there is an acceleration of the rate of change as well. So it's like a compounding effect. So things are really moving fast Annie. So it's like as you said we don't know what's going to happen. So it's interesting because this you know this year I'm running the Odonological society which is 148 years old. So, at the last meeting we had, I brought up this point and I said, you know, we just celebrated our 148th year um anniversary and we co-founded, that's a nice segue cuz Parsons and I co-founded this AI society at the school and I was saying that it's interesting if we look at this society, the AI society that was just founded in 148 years, where is it going to be? It's just want to know unfathomable, right? Yeah, but hopefully using AI I'll be and we'll all be very much alive of us to see the evolution of we'll be up I'll be uploaded right to the you know cyerspace going to be officially named James Cameron's successor at that point Dr. So we'll go from end to we'll have like AI podcast versions of us still playing it'll be that's exactly true but it's like the whole idea of education and the avatar thing that we one of the highlights again for the audience who had who didn't get a chance to to attend and hopefully will attend next time is uh one of the keynotes Dr. Cotch uh in Dr. keynote, he he did this amazing kind of um video showing himself uh you know, having not recorded this thing basically representing his his look and his sound and everything that represents who he is in an avatar format speaking in another language and it showed a few things. It showed number one that we can do that using generative AI and using uh basically some of the amazing image and video processing power we have with AI. But also he was trying to prove a point and that point was you know in the first example he showed the video the sound of the words he was saying and the actual movement of his lips weren't matched. So this was to the point of don't fall asleep at the wheel. Don't delegate something you know don't let AI create something and then don't even take a look at it. Monitor it. And it made me think about something that Dr. Nasseh has been talking about actually now for about a year or so which is how can we use this in education but I I see this as like for folks like you Dean Giannobile Dr. Nasseh Dr. Koch who spend so much of your time teaching and educating the youth and your time is so limited this could be an amazing way to sort of clone and replicate you guys to be able to get more insights. How does this make you feel? Is this something you'd be interested in doing? Would you would you be open way?
That's what I always say. If you ever see me online doing anything in any compromising position, it's not me. Just say my exactly. But yeah, even his second point, the that first video was maybe a four months difference between the second and the second one was that was almost scary. It was like, wow, this is as if you could have convinced me he had just came in and filled. By the way, this podcast is being translated to different languages by AI. Wow. So, I was just going to say I like that. I mean, we can do it in Spanish if you want but you know so so yeah I guess the question is would you is that something as educators and and leaders in the space do you think that is that where you see the field of go going in education AI like maybe kind of allowing experts or subject matter experts to kind of like use AI to the effect of educating our students. Is that something you thought about and maybe maybe not.
Yeah, I think it's going to be very helpful in supplemental learning and tutoring. And so this is already occurring at Harvard Business School as a part of the Harvard Innovation Labs. So they actually have these tutors that come in as a part of continuing education courses. after they have listened, it it could be you or I, you know, talking if they've got enough information about us, they can fairly well predict how you're going to respond to certain case scenarios. And so it can be very helpful in the pre, you know, the the pre-planning and, you know, working out different scenarios of challenges. and then when you actually have the meeting coming together to talk to the professor and to hear directly from them. Uh but they've shown some examples uh from the innovation labs and along those lines I might give an example that was presented in one of the workshops and maybe you've seen this by uh Dr. Jennifer Chen Heroe Oyama and Peter Greco and they gave a workshop on cavity preparations. I don't know if you had a chance to see that but it shows how you can look at all the dimensions of a cavity preparation and there are rubrics right on how how much occlusal reduction the shamfer and all these aspects and then scanning it through an optical image and then coming up with those that information that you could do in the middle of the night working on your your crown prep then the next day you have the faculty member to but there's a report that would be generated so I think this is where we'll be headed. And again, you know, to Annie's initial point, we do have to have the guard rails on that. We just have to have transparency on how this information is being used and transmitted because again, it's still we don't have robots completely doing everything on our preparations and endodonic fills, but you're the one who's responsible. Sure. As the clinician. Yeah. Yeah.
You know, it's interesting because outside of the the dental realm, I'm I'm I'm not intimidated by AI in terms of having a robot knock on my door and take me away, stuff like that. What I what I am concerned about is, you know, uh especially as people age, it's a natural part of the aging progression. You know, people are very concerned about protecting things. And when we think of how many people have hacked into major companies, have surpassed security and when that's just with computing. And now to me, I mean, you guys, I'm sure would feel the same way. AI offers so much more capability than just merely computing. Where how how do we protect things when you have somebody who is as equally smart as your people but doesn't necessarily share the same motives?
Yeah. Yeah. It's a good question. And when we had our regulatory session, and I'd be interested to hear your feedback. It's almost like with new intellectual property and you have patent attorneys working on your new invention. Now we have uh many people within the legal profession and we had an expert in that particular domain on how do we protect this and it's so challenging right now this landscape and it's going so fast.
I mean, that's insane. Exactly. I mean, there's no question there has to be some level of regulation because as you said Annie you know we now know these AI agents and bots can do all kinds of nefarious things right and soon enough if you get quantum computing for some reason somehow and it becomes you know uh available at at mass level then then it's a lot of bets are going to be off because it could do so many calculus and crack so many passwords but uh I think maybe I mean cyber security has always been about this stuff so you know I'm sure there are cyber security firms that are focusing on the interaction and kind of a um kind of a transposition of AI and its capability on the current existing infrastructure of the net. And I'm sure they will have to be and that's probably where the government comes in is where they're going to probably have to allow certain um I know allowances or something like that so that they can make sure that AI doesn't get out of hand because it it certainly can.
Ali, I have a quick question for you. All these things, who's paying for this? That's a good question. Anybody running for Well, there's an act actually under the panel. Dr. Cotch, to your point, I was listening to CNBC this morning before the clinic just to get a little dose of something different than dentistry. The future in the middle Yeah. They're in the middle of uh lobbying, you know, some of like the big tech companies for uh, you know, a set of bylaws to rush it through Congress because of this exact topic. And you know, I mean, I think just when we focus in on our industry and our profession, one of the benefits of a symposium like this is to, you know, really educate people who either are really bullish on it, like I think everyone here is highly bullish and into the this the notion and and focus on AI, but there's also people who want to learn and people people who are on the other end of the spectrum. So our job hopefully within our ecosystem and our school is to educate people on what's happening. Now then it becomes a question of bad actors versus good actors that one of our keynotes talked about and then it becomes a derivative of recruiting. Can you get enough smart people to be on the good side to fight against the bad actors, you know, some of these cyber criminals and whatnot? And in our space, it's um I think there's enough good happening across research, across um you know, diagnostics and treatment planning that I think people are excited enough. But one thing that I'm curious about is the guidelines that we need to set up. The question becomes, is this is this something that should be a federal mandate? It should be a state-by-st state, you know, local organization mandate or should it be international body?
International body. I know. What do you think about in dentistry? How could that look given the way that our organizational dentistry is structured? What what would it look like?
I mean, to me, doesn't it feel like we're I mean, dentistry is just we are essentially surfers. We're riding the wave. We have no control over the wave itself. So, the real question here is and what Annie's alluding to is the issue of how do we manage the wave? We're just one of many surfers if you will right so our in our profession we're just using the technology but the question is far deeper the fundamental aspects of technology and how that affects dentistry as well as every other field and that is really the main question that has to be addressed I think once that foundational stuff is laid down then dentistry will like every other field probably be safer but I don't know unless are you aware of anything will that is happening at the um you know ADA or any of these things that are
yeah, I think all the authoritative bodies within dentistry they're all trying to wrap their arms around it right now so American Dental Association it was even probably 5 years ago that they were putting together guidelines on how oral maxacial radiology programs need to be including AI as a part of their training um I I've been at two different international conferences recently on they were consensus conferences and they both included AI components with them. These were within periodontology. I'm a periodontist but we've also seen with the world health organization they have a network and at the symposium we had several members of this network and really trying to come to grips with how it's going to be used. So, I agree with what you're saying is that we're still trying to figure it all out and we're really relying heavily on many of these other regulatory experts, but uh the genie's out of the bottle, if you will, with a lot of these technologies and how they're going to be used appropriately.
Yeah. But something very kind of go back to my question. Oh, yeah. Please. Who's paying for this? How does this get how does this get subsidized?
organizations and obviously at some level it should be the government as a whole the federal government and the reason I say that is like if if you follow the economics of this and that's a whole another discussion I mean partially you should you know six or seven companies comprising 37% of the Dow that's kind of an interesting that's never happened before so I think what what's crazy is that there's mid-level investment houses in New York and they're now floating out trillion dollars and to me will I'm kind of like I remember billionaires what happened to that decade We gone from the B world to the T world, right? And so um this is what happens roene when you age. You start trying to go back and you know circumspectly look at things and rethink things. I'm trying to think like this is all wonderful but at the end of the day it's kind of a zero sum game. Who's funding this? Who's paying for this? Where where does that factor into the entire thing?
Yes. Well, I I I don't have a a good answer for you on that one. Annie, um I will say uh Ali and I we were at a a clinical innovation advisory board meeting and parcel you presented at it and an individual who was leading the biomed accelerator at Harvard University was talking about the investment in biomed research. It's really biotechnology is there's been a a major drop in the investment 50% investment really going into AI. So there's so much investment into this area and so to really understand who is going to be have paying and regulating all of this uh AI technology we need someone from the investment uh banking group to join us as well. The data center thing to me is is incredible. I mean the development of these uh ideas and uh things is just like capitalism. So there's modules of people using private investment to create the tools, but the overseeing of the way these things interact to create potential security threats for everyone. It's kind of like the same thing maybe in the defense industry, right? I mean, you got private companies creating all kinds of uh things that could be potentially nefarious and there is some level of oversight at the federal level level with these things except that now it's moving from that into technology.
Yeah. Yeah. What were you going to say earlier?
I was going to say something interesting and something great that I think that Harvard did especially for their students is they offered us an AI sandbox. And so we have a space where we have access to claw GPT4 kind of all those big engines that you could be you know giving your data to. But they said no here's a safer space where you can kind of put your guys' stuff here for now till we have a better handle on like where everything's at. It doesn't share with them. Right. Because I mean, there's always the big question of where is all my information going and that's you know and I mean a great thing they said at the the talk was if you're using Google Chrome you don't care about your privacy. I thought that was really funny but it's true we don't think a lot about this stuff but to have institutions that are willing to give you a safer place to kind of keep things for now till we all have a better read on where we're at is honestly like a really nice thing. And I think yeah to some degree it does fall in the institutional level and I can imagine it beyond and I'm sure there's eventual collaboration between Rowen that brings up a great question from I think is a sense that Ali addressed it earlier in terms of a generational thing but all right everything's coming well what happens to the provider who's 55 years old how how do they become uh more educated about the benefits and the applications and the use of AI I'm I I threw that out not as a rhetorical question that will be the job of the companies I guess Annie toh provide the education to bring people up to par. I mean, they may or may not. And if they don't, then I'm sure there would be third parties that would take up that task of educating people about AI. There certainly are a lot of uh places people can take AI courses and so on. But maybe one that would be and that's actually an interesting thing. Maybe that could be Hey, there we go. That's an idea for create a module to educate dentists about the uses of AI. Think about don't take your eyes off the ball. I think that's that's incredibly sage advice.
Yeah. I mean, I think it's it's obvious that it's education, coming up with continuing education programs that can really lift everyone up. And I think in larger organizations, even the development of an AI health officer, somebody either within a hospital setting, you know, many people are hiring digital gurus with digital workflow in the dental office, someone who is also fasile with AI because it's so difficult to keep up with it. I mean this is this is a position we're even looking at at the you know, the Harvard School of Dental Medicine. How can we have digital experts who are on top of this as much as possible? You know, we're trying to inculcate you know within the community that we all need to be involved with it but as you're saying the 55 plus cohort you know how can we get them up to up to speed on this. It goes back to the poll that you said about 70 some% didn't trust that it's you know usually we don't trust things we don't know. Yeah. And and then that's the key is how do we get to learn more about it?
Your point is so important because this happened like during the com boom I'm sure like you know the advent of the CTO chief technology officer whose domain domain whose main job function was software development and then it became chief innovation officer which is bridging you know software to the next level and and now we're at the chief AI officer point. Um but one of the things you Dr. Cut you said you know how do we get the word out? How do we educate this cohort maybe age specific? Well, it's events like this I think is important. You have to have cutting edge institutions lead the charge. But also on that topic one thing that I think one of the great speakers at the symposium and I'll maybe do a quick shout out Dr. Schwendiki one of the top researchers in the world uh in this subject um super super knowledgeable. He talked about how I and he's been talking about this idea of like a benchmark of a ground truth. So, you know, when a company walks into a practice and is trying to sell their imaging software and it says it performs 87% uh accurately, what does that compare to sort of the average general dentist? Is it, you know, how is that data being trained? So, the clinicians, the doctors don't really know what they're being sold. And so what he's been calling for as a call to action internationally is a benchmark uh basically this this data set this consortium of data where every product or company that deploys a product can actually compare against this data so they can be convinced or empowered with the information that it performs better than this standard. So Dr. Jones who's never used AI in their life and she's 63 years old um can be can be sort of educated on on how these things are performing relative to some sort of standard status quo right right right like but doesn't that create a challenge because in dentistry we are lacking that information even among natural intelligence so much among you know artificial intelligence I mean there is the standards of care are questionable as you see from even one state to another they may change in our field with endo for example, you know, what the kind of practice you have in the coasts is very different than the kind of practice you have in the uh, you know, central part of the US. So that might you know, this could be an opportunity for war and chaos or it could be an opportunity to bring people together to kind of agree potentially at some point as to what constitutes standard of care, right? I mean, crown preps different philosophies, different ideas, different people do different things, right? I mean, whatever we train the AI with is going to essentially be the future of the truth, right? I mean, cuz truth is unknown and we're all seeking it. But the question is those people who have access to training AI are going to control the narrative and are going to essentially redefine the profession in many ways because they're the ones who are going to essentially, you know, based on the training and based on their own manipulation of the data are going to create the reality that all of us will have to embrace whether we want it or not.
Wow. Yeah. Oh, sorry. Yeah. Go ahead. This is for Will. How do you as the dean manage to integrate AI into the basic curriculum that we've had for many years? That it's got to be an incredible undertaking, right?
You know, I I feel very fortunate we have a strong partnership with Harvard Medical School. So Rowen, you've gotten some of this education that's really bundled into Harvard Medical School, which really is a leading institution to really help at the very early stages of a student's training career. And then we are doing our best to keep up. We can't say that we're, you know, so so far ahead of any other place because it is constantly changing. I was just chatting with uh Sang Park. she's uh le co-leading a task force on AI and dental education and so we've been looking to provide support for projects within the faculty to develop educational materials. So I gave the example on the preclinical you know models using those but it is it is a challenge and we really need to help lift the boats if you will in terms of getting more and more people educated within AI. It's all about education and providing investment in our faculty uh to get those skill sets and then parsea like this AI society you know getting it reciprocally from the student community uh and the students inspire us and the students are discovering things every day and so that's where we're fortunate to have you know our our colleagues our budding colleagues and you'll be our you know our student doctors to really help building that culture of AI is crucial to dental education and the advancement of our profession. And so, you know, we're not there yet, but it's a work in progress. And I think now that we everyone at the school knows you got to be involved with AI at some level and how are you going to build it into your education, be it clinical or preclinical dactic. That's like really the the the crux
of the issue, right? I mean, all destinations start with a vision. It's your vision and leadership of the school is aligned with that idea of incorporating AI to the extent possible in dentistry. Correct. Right. Absolutely.
So that's you knowing the rest of we can say maybe from our perspective that that's amazing to hear and I think that's uh just it gives a lot of confidence and I think excitement to the maybe the younger generation. Um because I think we feel strongly that this is going to be sort of such a dominant force just like the advent of the internet. This is even bigger than that as you said Dr. Nasa. So to hear you know Dean Giannobile, Dr. Nasseh, Dr. Koch you guys say that give us that empowerment I think it means a lot right.
It's huge. I mean even seeing it I mean I'm seeing it in the classroom. So it's really interesting concept. We'll prepare for a course, we'll take the course and then we'll consolidate parts of the course. And part of the consolidation that we were seeing was you would be presented a case. Case would be patient presents blah blah blah. You can then interact. They created an AI tool Harvard specific as a consolidation exercise. You ask the patient questions like past medical history, chief concern, what are you feeling? It'll answer it for you. Then you'll move in to order your tests. It'll tell you, okay, this is available, this is not available. Then it'll give you all the results. You'll then piece the puzzle together. You'll give it your diagnosis. it'll give you a rating out of six. Such a cool concept to be sitting down, taking the information that we're learning in school and having a chance to practice asking questions that you can't really do without someone else there to to talk back at you and get that in real time. You know, you can't you know how sometimes you can like cheat on the sheet and like try and check the you can't get that. you really got to work your way through some of the questions and and just to see that and to be in the space where I'm starting to practice using the tools, starting to see how they could be used in action, but also to be at an institution where I think that we don't just say yes. And so seeing my classmates, you know, question the sustainability of AI, seeing my classmates question how much we even should be relying on it because we look at the faculty ahead of us and I mean we look up to you guys incredibly and we want to, you know, be able to get there also, you know, with our efforts in our own hands. And so it's very important, I think, to find that balance, too. And so, you know, I think it's it's very unique position to be to feel the support which is so wonderful and and to feel that healthy level of skepticism and and confidence and everything, but also just to see the things that you're all building between the AI symposium, the HSTM AI society.
How are you guys feeling now? You know, bringing these these new groups, this this new um you know, big event kind of moving forward. What do you see it changing for the way we're thinking about things? anything you guys are looking forward to coming up with with any of it?
I think it's the same kind of idea as what Will's vision was about bringing people together. Yeah. Is in some ways in all of these efforts, the idea is the understanding that no one person can know it all and that the collective together as a network could do a lot better than each individual. So the more people the better. So creating a society, creating a symposium, putting together groups of people together, they could exchange ideas, that's really the essence of um you know everybody working together to do better together. It's just a cooperative thing that is I think the basis of these efforts.
So what do you think Parson? What was your vision?
Yeah, I mean I think um it's just the beginning and I think you need a little bit of uh fuel in the fire to get these uh ideas deployed and underway and I think we've been able to sort of start doing that with your leadership and uh the symposium. I know today we're really talking about how the symposium was and recapping and I think to to put in a nutshell it was uh it was a hell of a weekend for anybody who's bullish on AI and I would write you know as the next one comes around in a couple years we'll we'll do this again but my call to action to anybody listening is try to learn more about it try to watch as many videos about it try to talk to as many video try to talk to as many people as possible about this idea because I think as you said Dean Genobi it's here to say stay And there's pretty much no turning back. Right. Absolutely. And I think it's it's incumbent on all of us to kind of contribute and share our ideas. And maybe Annie's idea of putting together a little uh thing to create and improve, you know, AI literacy for dentists is not a bad idea. I think it's important. We're all pretty connected within the school, but yeah, what about beyond? Definitely.
I have a question and again this is probably most directed towards Will and that is about partnerships. I think one of the wonderful things about doing this is doing this within a university setting of course. Have has has the dental school established some partnerships obviously I would assume with the med school but have you been involved with HBS? Uh has anything come out of them or out of the public health school will?
Yes I mean I absolutely I mean as we can expand the network as has been just discussed with other schools uh the business school. So I I gave an example of I I had the opportunity to sit in on a Harvard business school course on dental AI and these business school students knew they came in their experiences were getting their wisdom teeth out and having orthodontic therapy. Uh but you know Dr. Lani sat down and put together the the questions on how AI could be incorporated and all of these students they were raising their hands really getting involved in it and so education and the participatory part of it. So, we haven't talked so much about the patient, but when you talk about precision or personalized medicine, it's how we can get more and more people engaged on their oral health and how AI can really help support that. And so, engineering, you know, other schools, uh, educational programs are important, but industry, the governmental sector, and then this patient community, how do we better engage patients? And so I think patients now are going to be coming even more prepared. So when they come into the office, they're not going to just look at Dr. Google, but they will have this visual, you know, they'll have this avatar of a patient companion who's really going to help them based off of their basic or their information to help them in their own clinical decision making because soon enough they're going to be really partnering with these individuals to help them for their healthcare delivery.
Is it is it fair to say, will that you anticipated increasing the access to care?
Yes, it should abs that's you know um Hazen Alani is a faculty member in our oral health um policy epidemiology group and she has really engaged in that we uh wrote a paper on harnessing AI to reduce health disparities. uh and in that in that paper the concept is we can level the playing field not only with virtual access but with education with clinical information and so I think this creates a lot of exciting opportunities to engage you know all all you know components of the health care sector and education to advance patient care.
Do you guys feel that AI, at least in the way it is now, is already there? It feels to me like it's not. I just wanted to test it and yesterday I put up an X-ray of a case that I had done an APO on on a central incizer just said, you know, describe what you're seeing, make a diagnosis, see what's happening. Yeah. And it gave me a very very confident and detailed diagnosis and analysis of the X-ray and everything. And it was all wrong.
So, well, it's hard, you know, this talk about expert annotation, right? you know, I still need it. What I'm trying to say is that, you know, so some I knew that it's wrong, but the pro and it was a, you know, it was a CHP 5.1, which is like the latest model. Was it completely wrong or was it wrong? It was completely wrong. Like, so it was like every tooth that was on the X-ray had a previous root canal and this one that I had done an ape on had a post, which it didn't have a post, and that it had an infection, and it gave a complete, you know, thing on diagnosis cuz it was an apectomy. So there was like a raducency but that wasn't due to the uh uh there was an infection it was just the hole but but so it was so to me I would just wanted to see what it's like in terms of his analysis and and so if patients are going to it and it exerts the same level of confidence in terms of making these diagnostics that could be scary right so I don't know I mean you made a great point to bring this together which is like they are going to be so much more empowered with knowledge before they come in and you know maybe before certain patient demographics would have access to these information and data points and and little little things that some other patients wouldn't because of their socioeconomic status or education but now it's at their fingertips but then the question becomes you know what is the information that they're getting I have to say some of these like you know public models that we use like the claws and the anthropics and the geminis and the chat GPTs they're building themselves based on the data that they're collecting billions and billions of times a day based on the things that people are talking about and feeding it. Dentistry, I think, is still one of those domain specific topics that hasn't been really used. Yeah. So I think it's good because it provides a value for the businesses and commercial entities that are focused on that and I won't name them but there are a lot of great ones that are doing that provide their own value in itself and I mean I but I have this hunch and I hate to say but I have this hunch that these companies these large large companies AI companies probably within the next 12 to 18 months will be outperforming some of the dental specific or let's say nephrology specific or ophthalmology specific AI companies because the access to their data will be so much more robust and fast that it will kind of start to outpace the specific companies focus on specific domains. But I have to end with this this idea that you know we always compare what AI gives us as an output to what an expert is expert is thinking. So when we talk about perio and endo, we're have we're the cream of the crop right here, right? It's hard to for them to ever I think in the next let's say in the next 6 12 18 months get to the point of knowledge that we're talking about top-of-the-line experts in this in their spaces thinking and and and and perceiving this information. So it's a tall task to expect it to maybe look at things and endo like you or perio like you Dean Giannobile, right? I mean, it's it's not so easy, but No, it won't be easy. I just probably would have preferred a little bit of humility in to say that look, I'm going to give you a diagnosis, but I could be wrong. It just gave a beautifully confident diagnosis with the full thing and this is what's going on. This is what you see on the X-ray. This is these are the findings and this is the uh the full diagnosis. And this is the recommended treatment, too. It kind of made it cuz I asked for it to to do do all of that stuff and it just gave it without saying that, you know, well, you know, I could be wrong.
Oh, that's my that's my biggest issue is that if I'm going to ask it a question, I hate when I ask it and I'm like, if you don't know, it's okay. Just tell me you don't know. Don't make it up. Like, why are we creating something? Yes. No, I mean for patients they, you know, they the buyer beware, they have to be prepared to go to the expert to get the interpretation of the the diagnostic information. And so, you know, Ali, for you to give this example, you know, what's part of the definition of an expert is someone who's made almost every mistake Yeah. possible and they've learned from those mistakes. And so with AI, these models are, you know, they they've got their training data, but it's not necessarily curated for every different domain, right? And so they will continue to improve over time. Uh, but I think a lot of these rubrics will be for certain specific areas. It'll be more cut and dry on dental carries there. It that's still not even there yet. You talked about the ground truth getting to that but you will still need the specialist. You will still need the visual observation uh to you know do both the diagnosis the treatment plan especially of some of these complex cases. So I guess the good news for Ruin and Para coming out of the school uh soon in the first is that you guys still going to have a job and be needed for foreseeable future. Maybe better in that way. dentistry I think is actually fairly safe from AI for the foreseeable future. I'm sure during your I mean I feel like a lot of experts always make these predictions that there is going to be or a cure for this or that like within the next 10 years. So much of that is driven purely by um you know by just the economics of things. Yeah, I remember to just give a quick example before we end that and I I took a course at the main campus in 94 by Glimshire who was you know the guy who just developed uh who discovered the process of oification. I remember we had somebody come in and give a talk to us at the time. Uh so the course was on vertebrae skeleton and like bone physiology all this stuff and the person who gave gave a talk was talking about stem cells which were really big at the time right in '94 1994 and he said that specifically that knee uh pro all these people that have knee issues and stuff like that and you know cartilage issues within the next five to seven years we're going to have stem cells that are going to replace the whole gone cartilage. just going to come back and this and that and everybody in the class was awesome. This is going to be so great new dawn on medicine with these stem cells. Well, you know, just the last that was there for my knees. I want to be I need some of these stem cells. And I was like, you know, that was supposed 94 they said 5 to seven years. Are we there yet? They're like, oh, forget about that. That's not happening. So, it's true. You know, it's going to be a little bit of time.
All right, guys. So I think it's uh it's very fruitful the overall summary. You guys want to put that together? What do what do we learn from today?
Yeah, I think um you know today's conversation just recapping the uh second international AI symposium. Dean Giannobile, thank you for putting it together. Um as people who are highly bullish on this and and focus maybe half of my career on this, I'd say is uh it was exceptional and I really look forward to the next one. So, I I I'm hoping that we can have a a repeat of what we had this past weekend.
Oh, it'll be better. Don't forget looking forward to it. Exponentially better. Yes, we're exponentially. Absolutely. Okay. Terrific. Thank you so much, Dean Giannobile. My pleasure. Thank you so much. And holiday greetings to everybody and we'll see you in the next podcast. Yes. Thank you. All right. Terrific. That was great. Great job. That was a great conversation, guys. Thank you so much for answering my question. Where stories unfold.