164: What Happens to Human Expertise When AI Takes Over in Medicine

Will AI make doctors and specialists less skilled—or even replace them?

That’s the question I explore in this episode of DigiPath Digest #29. As someone working where AI meets digital pathology, I’m both excited and cautious about how automation shapes our skills and professional identity.

In this episode, I discuss two studies that ask tough questions about AI, expertise, and the future of medicine.

What I Talk About:

1️⃣ Endoscopist Deskilling After AI Exposure (Lancet, 2025)
A multicenter Polish study found that after frequent AI-assisted colonoscopy use, endoscopists’ adenoma detection rate dropped by ~6% when performing procedures without AI. It suggests overreliance on automation can subtly dull vigilance.

It reminded me of how we depend on GPS instead of remembering routes—or how driving an automatic car changes focus. Could medicine be facing a similar shift?

2️⃣ “Will My Expertise Be Devalued by Machines?” (Bangladesh, 2024)
Healthcare professionals shared concerns about:

• Job security and evolving roles 💼
• Ethics, accountability, and trust ⚖️
• Losing the human touch ❤️
• The need for AI training and oversight 📚

AI adoption isn’t just technical—it’s behavioral, cultural, and deeply human.

My Take:

I see AI as a partner, not a threat. I use it every day for research and content, but I never outsource judgment. AI can boost efficiency—but only if we stay curious, critical, and engaged.

We can’t let convenience replace competence. AI should augment our expertise, not erode it.

🌍 PathVision 2025 — Sept 5–7, 2025

I’m also thrilled to share that I’ll be livestreaming PathVision 2025 from September 5–7, 2025, on LinkedIn and YouTube! 🎥

This year’s conference is packed with innovations in AI, digital pathology, and cancer diagnostics. I’ll bring you live insights, interviews, and key takeaways from the sessions—so mark your calendars and tune in!

🧩 Key Takeaways

• Continuous AI use may lower independent performance.
• Professionals worry about trust, ethics, and losing skill.
• The goal isn’t to resist AI—but to use it critically and consciously.
• The best outcomes happen when AI and human expertise work together.

🕒 Episode Highlights

• 00:00–06:14 | Welcome & PathVision preview
• 06:14–17:46 | AI deskilling question
• 08:27–14:05 | Colonoscopy study results
• 27:11–38:10 | Healthcare workers’ AI concerns
• 43:59–51:05 | Reflections & responsible AI use
• 51:05–52:55 | Closing thoughts + PathVision invite

🧭 Mentioned

• Lancet Gastroenterology & Hepatology (2025): “Endoscopist deskilling risk after AI exposure”
• Bangladesh Study (2024): “Will my training be devalued by machines?”
• My Book: Digital Pathology 101 (Updated Edition Coming Soon)
• Event: PathVision 2025 – Sept 5–7, 2025 (Streaming Live!)

Thanks for listening to DigiPath Digest #29! I hope it inspires you to think critically about how we can embrace AI without losing what makes us human.

And don’t miss PathVision 2025 (Sept 5–7, 2025)—I’ll be streaming it live for three days of insights, innovation, and community. Let’s keep learning and leading the future of digital pathology tog

Transcript

00:00:00 - 00:01:15
Aleks: Good morning. Welcome my trailblazers. Welcome to the 29th Digipath Digest where we explore the uh literature about digital pathology and medical AI. When you are here, when you join, let me know where you're tuning in from. And the chat is working. We're back to our Okay, I see you're joining. Whenever you're here, let me know where you're tuning in from. I'm in Pennsylvania 6:00 am back from uh Park City. So, let me give you an update while you settle into the live stream. We have coffee. The Polish cup.


00:00:38 - 00:01:59
We need coffee at 6:00 a.m. I mean, I don't if you do. I do. I just came back on I don't know when. Beginning of the week from the Lumea here. Luma digital diagnostic summit. Who might have tuned in into the live stream, and they gave this beautiful mug or jug and I put already the digital pathology trailblazer sticker on it. I have the book to later offer you the book. I was supposed to work on it this month and finish the new version. And we have Scott from Atlanta. Scott, you're a regular on these live streams. Thank you


00:01:21 - 00:02:37
so much for joining. So, you're amazing. H and I was preparing for this one because I skipped in September because I was supposed to be working on the book. We did two live streams where I was doing stuff on the book and then when I was not accountable to you, I stopped working on the book. Not great. But uh what I want to tell you is that I was so excited to join you here live h and because you're amazing and there about you being amazing that I want to tell you I was city h together with imagining fit


00:02:00 - 00:03:12
and we're going to go to Path Visions together as well. I'm going to tell you about my path vision's appearance in a second. But I was streaming there and we were talking about you know all all things digital pathology and one of the things was that um there is an image-based test stratification test um for patients uh by Astroenica that assesses the membrane versus cytoplasmic staining of a specific IHC marker tropto and I was like stumbling upon like which marker is that and sure enough in the


00:02:36 - 00:03:42
second there was exactly what the name of the marker was, what the name of the test was from a trailblazer. And that's why a huge shout out to you. Um, and obviously as I like think, okay, how to create better content, how to create more content, how to create more podcasts for you. Um, I was thinking of different formats and the one that I enjoyed the most are these live streams. later my editor takes out all my rambling and all the audience interactions and they become podcasts. Um, so yeah, I'm going to be


00:03:09 - 00:04:11
doing more of that and I'm going to be doing that from Path Visions. So, uh, at Path Visions, I'm going to be there as a media partner. So, thank you so much digital pathology association for uh, joining forces and being a media partner and I'm going to be streaming from there. So, if you see me walking around with my camera, if you see me walking or streaming, you can just wave and join me. I would love that. Like, don't be intimidated. That's what I do. But I do it for you. So, if you want to join me


00:03:40 - 00:04:43
on this side of the camera whenever you see me around, you are more than welcome. Just, you know, wave and it's going to be our signal that you join. Or maybe we should have a special wave or anything. just let me know like I'm there because I want to meet you as well. And we're going to also schedule a lunch when uh you can join me. And obviously I have to do it today because I'm flying on Sunday and you're probably already like prepared and booked for dinners and everything. Uh but I'm going


00:04:11 - 00:05:17
to send it out hopefully today in an email when we're meeting. Probably Monday. I'm thinking lunch for Monday in some space where you can just join me. I'm just going to be eating there and whoever comes comes and we can hang out together. Uh I would love uh you to join me. I'm going to be vlogging of course, streaming as I said and of course supporting the sponsors, the digital pathology place sponsors. Um so I'm going to tell you uh later at which booth uh you will have the highest


00:04:45 - 00:06:00
probability of finding me because we're going to be recording a podcast with Mtuitive. They're gonna be uh social media uh posts created with Barco, with Grandium. Um probably I'm going to be hanging out around a Predia booth as well. Everybody who is a digital pathology place sponsor and definitely I want to meet them. I want to see them. Um and today's topic, let me know where you're tuning in from, but today's topic is oh my goodness. something that when I have not seen a


00:05:22 - 00:06:48
paper having so many press releases and being picked up by so many media sources. Let me cover to use some water from the Trailblazer Lumia cup. Amazing uh amazing summit by them and I visited their lab. There's going to be videos of course um and you can see it. So but going back to the topic which is is AI a threat to us in terms of us as healthcare professionals as basically professionals losing our skills and everybody somehow under the surface uh everybody or let let me just stick to me


00:06:14 - 00:07:22
uh me having used claude chat GPT and a lot of these uh different tools I'm kind of like pushing this to the back of my mind, but I'm actually asking myself, am I getting like lazy and stupid by using them? And I'm afraid that at least there is a partial yes to this question. So going back to the papers, let me share uh the first paper that we're going to be talking about. It's endoscopy's deskkilling risk after exposure to artificial intelligence in colonoscopy


00:06:48 - 00:07:59
and multicenter observational study. Um I'm going to put it big in a second. But what I want to talk about is like how many media outlets picked this paper up. The first time it's in Lancet. The first time I saw this paper, unfortunately my super pen doesn't work, so I'll have to do it with a mouse like the the drawing and stuff, but the first time I saw it was when I was covering the NCCN um summit on AI and healthcare and one of the panelists cited the Lancet paper and he just showed the Lancet um Lancet


00:07:23 - 00:08:22
cover and I'm like okay this skilling of endoscopies and it was part of discussion right are we going to get more skills because AI is going to help us with something or are we going to lose the skills on his cited this paper? And then I was talking to my husband, he's an MD, a clinical pathologist, and he says, "Oh, this study was published in Lancet. This is a study from Poland." And I'm like first shock, how do you know about this study? Uh I mean he helps me out at digital pathology. He's


00:07:54 - 00:09:01
actually at digital pathology place. he's my VP. But I'm like, hm, you're like, I'm on top of these things. So, I was surprised. Anyway, that's first thing. And second, Poland. And I'm like, Polish paper. And I didn't cover it in Digipath Digest. So, today we're going to be covering it. Let me make it big. And actually two papers. It's not the abstract today. Paper is uh Oh, where is my second paper? Okay, here's my second paper. Come on. The other one it's it has the


00:08:27 - 00:09:42
coolest title. I wonder if my years of training and expertise be will be devol devolulted by machine by machines. I'm like yeah I'm wondering as well probably it will. Um, and I'm going to give you obviously my perspective on this, but um, yeah, and the the conflict is AI is being adopted rapidly and it promises huge benefits and I'm taking advantage of it in terms of content creation tremendously. Like if people would take it away from me, I would be very sad uh, and a lot slower and I would have to


00:09:04 - 00:10:33
like use my brain. Um, it would be a lot more painful. Um so in this uh today we're going to be looking at two key perspectives. One is the death killing. Um and the other one is the concern of healthcare workers. Uh this one is about the concern it's it's from open um sorry see I don't have my pen open nursing. um and they applied psychological study methods to uh investigating okay how do healthcare workers feel about AI so let's start with the first one what's happening here h is there endoscopies


00:09:49 - 00:11:14
the skilling risk after exposure to artificial intelligence in colonoscopy a multic-enter observational study and this study is from Poland. Um, let me do this one. It is known, sorry, it is not known if continuous exposure to AI changes endoscopy's behavior when conducting colonoscopy. So, uh, they assess how endoscopies who regularly used AI performed colonoscopy when AI was not in use. And I like the study um, let me show you the methods. I like the study design very much because what they did is um they had four


00:10:31 - 00:11:45
endoscopy centers in Poland and when I looked where sorry of course it's not big it's small nobody's telling me in the chat that this thing is too small for you to see anything you must tell me in the chat because then half an hour passes and you don't even see what I'm highlighting so feel free to give me feedback in the chat but I'm excited for endoscopy centers is in Poland and one is in Gou Vapolski and goju veropolski h is here where is it does it say anyway it's very close to where I live I


00:11:08 - 00:12:25
go shopping there I used to go to I used to go to English learning school there and to musical school there so goof is close to my heart my sister went to high school there so uh they conducted a retrospective observational study um and u this was part of a bigger trial accept artificial intelligence in colonoscopy for cancer prevention trial and uh these centers introduced AI for poly detection at the end of 2021. So this was a a retrospective study after which colonoscopies had been randomly assigned


00:11:46 - 00:13:12
to be conducted with or colonoscopies with or without AI assistance according to the date of examination. Right? And they evaluated the quality of anoscopy colonoscopy by comparing two different faces. And this is uh what I like a lot. Maybe I can change the color here. Um because they took the data from 3 months before and 3 months after AI implementation. So we have a long period uh 3 months before they didn't know how to use it and this was like the without AI and it's going to be important because


00:12:29 - 00:13:55
uh it's kind of a concern like when measuring the baseline so this was the baseline um and then they learned AI and 3 months after uh implementation they took away the AI right and they included diagnostic colonoscopies and obviously they they like did it as standardized as they could and what they were measuring was the primary outcome was change in adenoma detection rate uh of standard nonass assisted colonoscopy before and after AI exposure. So the adenoma detection rate is kind of like a


00:13:16 - 00:14:55
measurement of success of colonoscopy. The higher this rate, the better they are in detecting these adenomas um which are supposed to be like pre precancerous lesions. And what happened in this paper was that um they did a lot of they and endoscopy they colonoscopied a lot of patients um and uh a lot is 1443 let me just spell out um and they underwent nonI assisted colonoscopy before 795 of them and after so both were non AI I assisted but the second period was after the doctors were using AI. So they were already you know


00:14:05 - 00:16:00
using this tool and it's a specific tool I think they they're mentioning it. So 648 was after the introduction of AI. Um and you know they give the data on the patients. Um and what happened here was there was a difference a decrease of minus6% of this an adenoma detection rate. So the uh ADR of standard colonoscopy decreased significantly from 28.4 4 uh before to 22.4. So we have this 6 decrease. Not great, but it gets worse even, you know. Uh and if you have comments, anything that uh already like sparks some kind of


00:15:02 - 00:16:16
um turmoil in your head, leave it a comment. Leave it uh something in the chat. Um, this live stream is not going to be a prescription. Oh, do this with AI or do that with AI and a pretty capable researcher conclusion. A lot more complex conclusion. So, obviously uh easy for me and easy for human mind is to have like five different steps that when you follow your going to be successful. You're not going to get that from this uh particular uh pot live stream. Uh but let's continue. Right.


00:15:40 - 00:17:13
So what did they conclude? They concluded that continuous exposure to AI might reduce the ADR. So might of obviously this was an observational study in my clinical trial. uh so it might reduce the AR of standard nonAI assisted colonoscopy suggesting a negative effect on on endoscopy's behavior um and why do they think it suggests and and what kind of behavior right so um that uh kind of supports this conclusion is when endoscopists were using AI there was this eye tracking movement done and


00:16:30 - 00:17:44
they were los they were using their eyes less than without AI and they basically and I can imagine this as any like uh computer aided diagnostic basically like you get something highlighted you go you confirm and if you have a confirmation this is your diagnosis so basically you're done um at least in pathology right you you have focus uh let's say I look at the slide and um you have a focus of neoplastic tissue, right? There is your diagnosis. You can already say this is cancer or what that is, right?


00:17:07 - 00:18:25
This inflammation, this is whatever. And you don't have to go and search for more of it. I mean, if you want to grade it and describe, uh that's fine, but you already found your answer. So, if you have an AI assistant pointing you to that place, then logically you're going to be using your eyes less. Uh so kind of you know stuff is logical but also concerning. So that's why no five um fivestep uh five-step framework to avoid getting less right five-step framework to avoid getting stupid with AI


00:17:46 - 00:19:11
and no we don't have that today but moving on so um they might uh this might suggest a negative effect on endoscopy's behavior and like if you Google this and dkilling like go now and Google this. You're going to find like so many media outlets in the medical space pick this paper up and I'm so proud it was in Poland. Let me show you a few few graphs from this one. I'm going to make it big. So very thoroughly like they characterized everything. They did the multivariate um they have one graph that


00:18:33 - 00:20:21
exactly says what it was. Yeah. logistic regression analysis to identify factors affectction rate and they um used like everything. Sex of the patient, age group of the patient, adequate bowel preparation and incomplete examination. Yes. No. sedation and also sex of the patient are like they're known to have a higher adonoma detection rate. Right? So they um they account right was it was it surveillance specialty? Uh was it physician surgeon? I wonder which one was in G and also the endoscopy sex and


00:19:27 - 00:20:58
what they found out and there is more more data needed on this that even though there were so they had 19 endoscopies in total and five of and the females dropped not by 6% by 15 when the AI was taken away. So obviously they are not making any conclusions but definitely there is a lot more like psychological behavioral uh research that needs to go into the interactions with AI of healthcare workers uh to figure out okay how do we mitigate um because okay let me let me show you a few more graphs of


00:20:15 - 00:21:22
this one and then we're going to move on to the next one and let me know if you have questions let me know if you have comments. Um, and and what are your thoughts? Are you like scared? Are you thinking, "Oh, we'll figure it out." Are you thinking, "Hm, let me read some more." Just give me a reaction in the chat. What are your thoughts to to this? Um, and also let me know where you're tuning in from and what time it is there. It's 6:20 here in Pennsylvania. So


00:20:48 - 00:22:23
um they did this also uh I made here this was the 28 original uh before AI introduction and then we had 22 more or less. Apologies for my dots. This was after AI introduction. And here is um they only have 15 here because they excluded I think the ones that had very little experience. But um look four of them were better and the rest was worse and somebody was like very bad. I would not be that person. It's anonymized. Nobody will know. Of course they know. Okay. Anything else I wanted to show you here


00:21:42 - 00:23:14
in this one. Yeah. So that that's the diskilling problem, right? That is one of our problems. But uh as we all know or maybe not but some of us know that uh as we try to change something in an institution that uh change management is difficult. Um and whenever a new technology is introduced there there is a level of resistance to it and uh I keep hearing for digital pathology AI and medicine there is always this question popping up and I thought it's going to like stop popping up but um it's not and


00:22:31 - 00:23:51
it is okay is AI going to take a medical professionals as scientists is you know contributing uh to healthcare maybe whatever right um is it going to take our jobs and there is like this saying or slogan no it's not going to take our jobs but uh if we don't know how to use AI then we are going to be replaced by people who know how to use AI although now after the this killing papers like hm those who know how to use AI when you take away the AI h it's going to be even worse Um but again food for thought here


00:23:12 - 00:24:47
about other technologies because before we dive in um this other paper is okay I was thinking about it and I'm like I know how to drive a stick. I'm from Poland. I know how to drive a manual manual. What is this? Gears in the car. And I drive it in Poland. And I'm thinking, am I and I spend now more time in the US. I have an automatic car here. And uh I was thinking, okay, am I worse doing this uh driving the stick in Poland? And when I focus, I'm not worse, but I'm used to not having to focus so much. And


00:23:58 - 00:25:26
uh let's keep it vague so that you know I don't want to disclose like how unfocused I am driving uh online. But let my friend does that. Let's let's talk about my friend instead. So my friend, she knows this how to drive a stick. H and she also um most of the time she spends driving an automatic car. And when she gets into the car in her country of origin where where she drives the steak, um she is used to being able to be a lot more distracted like um bad bad habits. I hope she stops doing this,


00:24:43 - 00:25:56
but picking the phone and doing some stuff on the phone, right, while driving is prohibited in Pol in her country. So, um she should not do that for sure and she's not going to be doing that because I'm going to talk to her. But, uh that like drinking stuff, anything. She knows that when when she drives an automatic car, she can afford to be less focused. like subconsciously, right? It doesn't impair the driving, but when you're driving the stick of course, it does impair the driving. So, that's one


00:25:19 - 00:26:32
thing. She didn't really get be, but she did get defocused. And another example, my friend relies on Google Maps. Uh when Google Maps was not a thing and GPS on the phone was not a thing, she would use a map. um she still knows how to map but she's a lot slower and she actually doesn't have to use a map. Like I remember uh okay let's let's go back to myself. I remember when I was trying to h get to one like some remote place h when I was doing my vet um we had to do every we had to do some kind of


00:25:56 - 00:26:59
internship and one was a slaughterhouse internship as a veterinarian. this is a mandatory thing and the slaughter house was like in the middle of nowhere and I remember like stopping at every crossing consulting this map and going picking up my next turn like I hope hope I never have to do it again or when I was and I would have the map and my mom was driving and I would have the map and I would like no we don't have to do it um I still know how to use the map um and Google maps kind of you have to


00:26:27 - 00:27:55
understand how it works um but these two examples came to my mind when I was thinking like am I really getting worse at the stuff that I have assistance with and in a way yes yes does it matter for these tasks when you quit this um 99% of the time so it doesn't really matter this one% of the time I don't know uh I will probably call somebody who has Google maps who can guide me or something. But enough ram to the paper. Our next paper is I my years of training and expertise will be devalued by machines because you know


00:27:11 - 00:28:22
the examples I gave is everyday convenience. Um not even as a professional driver do you have to worry whether it's a stick or not. um the car that you're supposed to use for your driving, then that's the car you're going to drive and you're going to learn whatever that is, right? But um the the thing that was interesting here, it was not just um my experience from a conference where people are asking me, oh, am I going to lose my job? uh which you know after you hear it several times you're


00:27:47 - 00:28:55
kind of like if you're in this space like if this is an area of your expertise you're think like why again the same question h but here it was um done with various healthcare professionals so this study is is from Bangladesh but the the concerns that they are addressing here are universal totally universal and I'm going to just interrupt for one second because I don't see too Any comments guys? Come on. Show some interactions in the chat. What do you think? Are you afraid? Uh are you


00:28:21 - 00:30:00
confident? Are you using AI? Are you not using AI? Let me know in the chat. What are your thoughts? Um okay. So they did this study design. In this study, a descriptive and exploratory research design was adopted to investigate healthcare workers and viewpoints considering the potential substitution of human medical professionals with AI. Right? So their action and um viewpoints. But what I wanted to highlight is that they were drawing upon that acceptance model, technology theory and sociotechnical systems theory. So these


00:29:10 - 00:30:26
are like um sociology and psychology theories h when it comes to introducing new tech. So there are apparently like investigated ways that people react to this. It's not specific to digital pathology. It's not specific to medical AI. It's any piece of technology that comes into our life. There are like certain psychological mechanisms that are being that are in play, right? So they investigated these mechanisms and what were the demographics of the people that were investigated. So um very cool


00:29:48 - 00:31:14
tables tables uh semistructured interviews and who was these people who are the so this is a nursing journal journal so I expect and there's going to be a lot of uh nurses so nurse nurse hospital administrator physician medical technician nurse physician technician nurse physician who else radiologist some radiologist pharmacist interesting some pharmacists [Music] physician and pharmacist medical. Okay, so this group uh and obviously gender, highest education, highest education, years of working. So


00:30:31 - 00:31:45
they also like had a very diverse group that they investigated and um they had structured interview focus group discussion outlines. They had like seven different things that they were addressing. So number one, how would you describe your current role in the healthcare industry? And can you share any experience or uh instances where you've encountered AI in your work? A and then what are your primary concerns and apprehensions? Right? So they were asking about like the negative stuff. This is what they


00:31:10 - 00:32:32
let me make this bigger for you. [Music] What's this? Sorry. Yeah. Okay, we should be fine. Um, okay. So, apprehensions regarding the potential replacement of medical professionals by AI. How do these concern relate to your daily work? In your opinion, what are the key ethical and societal considerations that need to be addressed when integrating AI into healthcare and how do these consideration influence your concerns? Uh can you provide examples of ways in which discussions about AI have influenced the dynamic and relationships


00:32:03 - 00:33:18
among healthcare workers, patients and within your healthcare institution? How do you envision the future of healthcare with increasing presence of AI? H what measures or strategies do you think should be in place to ensure a smooth integration? Uh see they they don't even ask like oh do you want to like get rid of it? No it integration smooth integration addresses healthcare workers concern. And what recommendation on or insights can you offer to policy makers, policy makers, healthcare organizations


00:32:40 - 00:34:07
and professionals? Right. Let me just drink some water. So, I'm going to make it smaller because there's a cool picture I want to show you. Okay. And here were the concerns, right? [Music] Ethical and moral dilemmas. Nothing new. Nothing new. We are all like um is it good? Is it bad? How is it going to affect? I have some comments from Thomas. Thank you, Thomas. I was counting on you. I thought if you are live here with us, you're going to write something. And I'm going to get to it in a second.


00:33:34 - 00:35:04
Let me just go through this image. Um yeah, so job security and economic concerns. Um very important. This had already effects in reg with AI assistance. Um it kind of like in the model of American healthcare where you have like fee for service the more service the more the fees I think in radiology became services and and that's a whole can of worms uh workforce role sorry role and definition trust and acceptance of AI healthcare population patient provider relationship quality of patient care Right. This one


00:34:19 - 00:36:05
is important because okay uh can you trust it? Uh and who's reliable? That's a big one here. Like a lot of um oh so this is like a combination of these two ethical and moral dilemmas. Then um okay what is our role then if like this AI system that does stuff for us? Um, are we then supposed to be like very techsavvy? Like how how does the healthcare worker worker profile change and yeah, let me just check the tables. We're going to go back to it. Um, but yeah, I want to I want to check Thomas's


00:35:13 - 00:36:52
comment is is long, so I'm going to just read it to you because it's an interesting I want to show myself interesting discussion point um is using AI and um find a very performing and sound specific specific task but not so intelligent globally. I like this perspective a lot [Music] and so simply it means we will work quite to very uh differently but at the end of the day we will be the guys signing in the diagnosis. So yeah that's for for pathologists. Um let me read the rest of it. So, so yeah, that is an


00:36:04 - 00:37:14
important point because uh all these tools are supposed to help us, right? And the deskilling experiment also like they didn't say, "Oh, we're now like going to take away this AI." The the um and that's like kind of a theme as well. The doctor with AI is doing better than the doctor without AI. H before learning AI, we just learned that that doctor without AI when they were already learning, sorry, working with AI, that's even worse than at the beginning. So just keep having them with


00:36:39 - 00:37:55
this AI, right? Like me with Google Maps or my friend with the um car the the stick versus automatic. So it's going to be a partnership. There's still so the liability questions, right? Sometimes it's very complex, but the person signing out is going to be still the main person responsible. But if you're using a tool that you are not involved in developing and like it fails, there is going to be a gray zone. Yeah. And and um the comments I get from Thomas is that there's a lot more more


00:37:17 - 00:39:01
intelligent, smart, critical review and decision making. That's going to be our work. Come on. Come on. And yeah, thank thank God it's Friday. Thank you, Thomas. I know it's like heavy topic for Friday uh morning or afternoon, but um yeah, concerns. What I want to say is uh in this picture that I just showed you um the policy makers no in some other well we're like um what will the policy makers do because uh that's ultimately their role to so uh the NCCN guidelines summit on


00:38:10 - 00:39:35
um AI in healthcare uh was supposed to basically address that. they were um different stakeholders including patients including uh you know everybody involved in uh cancer care um and they were discussing like you know they they were having questions they were discussing and then NCCN um policy experts are going to issue a report based on this and then the regulators the institutions the organizations ations are going to take it and think okay how can we guide the health care professional um the smarters


00:38:53 - 00:40:03
the smartest and we actually will have more expertise with AI I think so too like we're going to be better with the tool like obviously I like these tools right and I use them as much as I can uh but also in the back of my head there's this question okay I don't want to get I don't want to be stupid. So what do I need to do in another area of my life to keep my brain active and um thinking critically? So when I use these tools, I do engage with them. I very much critically read everything that


00:39:28 - 00:40:45
this the the chat GPT cla or I puts out because it produces nonsense. If you've seen me do one of the book live streams, oh my goodness, what nonsense it produces. And you then you ask it, oh did you like um invent this? Oh yes, I'm sorry. I fabricated this part of the book. And I'm like, yeah, I know because I have enough expertise uh to know that it's fabricated. But basically, like the critical engagement and uh with the tools, but not all the tools are textbased. Not all the tools are explainable like that. I


00:40:07 - 00:41:14
mean explainable. I don't know how this text was generated. I mean kind of know with like a transformer architecture but I can verify it because I can read and I can understand if it doesn't make sense then I'm not going to put it in my article right this is like easy cognitive check that I can do as a pathologist with a computer aed diagnostic tool that shows you some area uh with a specific diagnosis if you disagree with this diagnosis you have the expertise to say no right and if you


00:40:39 - 00:41:57
see like five nos in a row, uh you might go to whoever deployed the tool and say, "Hey, this tool is underperforming." So, we have this level of um potential intervention if we pay enough attention and we should because we sign out uh whatever our work is, right? It it has our signature. Be it pathology report, be it an article that you write, be it a book. And now it's time for the book promotion. Let me put this um let me give you a QR code that you can scan. So uh when you see this QR code, you will get the digital


00:41:19 - 00:42:32
version of this book and the book is digital pathology 101 for everybody who's uh starting their journey in digital pathology. It's very cool resource uh because you can get up to speed very fast, not that long. uh currently 120 132 pages. Um it's going to be longer because I'm updating it. But uh worry not because if you get uh the current digital version, you are already on the list of my email subscribers and you're going to get the new version as well with updated AI information and um


00:41:55 - 00:43:13
everything that happened since the first edition was published in 2023. and I am committing to publishing it before the end of 2025 was supposed to be this month. I'm actually this month so that you can get the um new version, but get the old version anyway because then you don't have to like sign up for anything else and you're already on my list. Um so yeah, let's do it. Ah because one of the uh one of the conclusions of the nursing paper of the paper uh will lose our job or is it like um one years of


00:42:34 - 00:44:10
training and expertise will be devol devalued by machines devalued I think I like butchered this word um one of the of the like recommendations and requests by the healthcare professionals was training what happened uh training was um the main thing that they were requesting and there's another comment by Thomas that these these tools the the these AI whatever it's going to be software assistance for robotics or whatever um they are tools that allow us to be more critical like a tool is not going to get


00:43:22 - 00:44:38
offend it when you like double check it and like critically like when I work with work with these um with Chad GPT right I'm not going to uh feel bad that I'm asking it for the seventh revision if I was working with a partner uh on a publication at some point I would be like let me write it myself or like stop interfering so much in this case like I am the guide right and I can have it the way I want I have it. So, um, lots of time in the olden days, it was lots of time to make great quality


00:43:59 - 00:45:08
manual analogical work and we ended up having less time to make high value work. I totally uh subscribe to that one. Um, and also if you have something like for me, I think it's invaluable because I want to put out a lot of stuff for you. you all like this is my mission to help you understand digital pathology and medical AI so that you can do your job better so you can feel more confident so you can be in the know and like you know after reading this 100 something page book you can go to your


00:44:33 - 00:46:03
meeting and contribute right h so to do that in the fast-paced environment I very much leverage these tools and thank you so much for this fantastic discussion Um if somebody ah I need to show you where I'm going to be at path visions. Let me let me share my screen with you. Path visions. Let's see. Okay. We have path vision. Shoot. Stuff happened here. Let me go back. Okay. I don't want to share my screen. Share. It's doing some funny stuff. Okay, I'm just going to show you who is going


00:45:27 - 00:47:03
to be there from my supporters and where are they're going to be. Okay, so Barco, no, let me do full screen. Sorry. Sorry for that. Ah, I know what I need to do. I need to take this to a different screen. That's not going to do any nonsense. Uh, let's do this. And thank you so much for scanning the code for the book. You're amazing. You like really are. I love doing these live streams. It's the 29th Digipath Digest that we have been doing. So, there is something to it. Uh, okay. Bear with me


00:46:15 - 00:47:39
cuz I do need to tell you where I'm going to be. Okay. So, fantastic, fantastic, fantastic. Here, Barco Booth 314. You're going to find me there for sure because we're going to be uh creating some social media content. And if you see any content from the Barco booth, uh, give it a like, give it a comment, say hi, whatever. just um let uh the algorithm take it further. Then who else is there? I think booth 400 a predia they are digital pathology sponsors so I'm going to be visiting them for sure.


00:47:00 - 00:48:14
Grandium my friends today the grandium the video about the new Grandium scanner is coming out. So, if you're going to Pet Visions and uh you can then watch the video if you're on my list because you're going to get it today and then go and check this super cute scanner. I already have it packed in the box, but uh I'm going to show it to you in this video. It's It scans four slides. I don't even remember. Four or five. I think it's four. Four or five. Okay. Now I have to check it because here I'm


00:47:36 - 00:49:08
spreading misinformation. Let me check. Let's see. One, two, three, four. I think it's four. Uh yes. Okay. So, there's going to be an unboxing video uh with the scanner, right? Then who else is my friend here? Oh, shout out to Luma because we just came back from their summit and it was a really good um event. So, if you are looking for something where you can connect with decision makers like in a pretty intimate setting, that would be the um the summit the the conference to go to. Very uh intimate like there was around


00:48:55 - 00:50:25
100 I don't know if even 150 people and we had these excursions. I don't know if you've seen my ATV video from last year and this year I will we were riding horses. So you connect with a pretty high level uh digital pathology and medical AI achievers. So um if you want to learn more about this booth 208 and Mtuitive we're going to be uh recording a podcast together. Uh this is booth 46. and Rash booth 401. So next to a predia my two sponsors are together so I don't have to like a lot techite we recorded a


00:49:42 - 00:51:06
podcast recently well recently in March at the US cup but they are my sponsors so I'm going to visiting them as well so yeah 400 oh the 400s 412 4001 400 something else anyway that's where I'm going to And if you see me walking with the camera, just say hi. Like you can tap me on the shoulder and say, "Hey, I'm a trailblazer." And then I will know who you are and you know who I am. And if you just are not listening to this, then yeah, let me just stop talking. Anyway, guys,


00:50:26 - 00:51:38
thank you so much for joining. Um, you are amazing. you make me want to keep doing this. Uh, and okay, the QR code, uh, I see some people had trouble, but I think it's working. So, um, also, if you, uh, if this code is not working, you can always go to digital pathologyplace.com and there's a button to, uh, get the book. If not for you, if you already have it or if you're like an advanced digital pathology trailblazer, just send it to someone. They will appreciate someone who you know is just starting


00:51:05 - 00:52:14
their journey or have expertise in a specific and specific, you know, area but want to learn about more areas and um if you are interested in learning more about AI, we have a code path AI pathology AI makeover. It's a new course that I uh created. It's already updated. So the um like some of the information from the course is going to be in this book. This one is paid. Um and also you can just check everything we have in the digital pathology store that is recently launched and obviously in the store you


00:51:39 - 00:52:46
can get the earrings. Earrings are not going to come before Path Visions but definitely before your next conference. So this is my favorite because this is the multi-ucleated giant cell. I don't know. I just love this cell. So beautiful, so ineffective because too much knowleated giant cells are the macrofasages that come together to eat something that's them. And um yeah, it's overwhelming. So they're not too effective and they just sit there. But I somehow like how how they look. And we


00:52:13 - 00:53:39
also have partilage. And we also have this pattern which is alian blue stain of colonic. So um probably that's going to be mostly female trailblazers. But uh if you want to give it as a gift, fantastic as well. What I asked the DPA did it last minute. I did I asked if um I can bring them to Path Visions and basically um have a part of the um revenue from this be a donation to DPA. Um let's see maybe they're going to be there. If you're there I hope to see you there as well. Thank you so much for joining


00:52:55 - 00:53:18
me. Stay as you are a true digital pathology trailblazer. I hope to see you at talk to you in the next episode. Thanks so much guys.