Beyond the Chatbot: Why AI in Healthcare Still Needs the Human Touch. Insights with Prof. Krishnan Ganapathy

Artificial intelligence is changing the way we think about healthcare, offering new ways to connect, diagnose, and support patients—especially through telemedicine. More people than ever are speaking to their doctor from home, getting advice online, and sharing data from their devices. It’s fast, it’s convenient, and it’s full of promise. But is it enough? Where does technology stop and real human insight begin? 

In a recent episode of Development Aid Dialogues, host Hisham Allam sat down with Professor Krishnan Ganapathy, a veteran neurosurgeon and one of the world’s leading voices on digital health. Together, they cut through the hype to talk honestly about what AI can—and can’t—do for medicine today. 

Ganapathy doesn’t shy away from the benefits. He’s seen firsthand how remote consultations and wearable gadgets make it possible to spot health issues early, save time, and reach people who might otherwise be left behind. He’s comfortable with the future—“A clinician who’s not AI literate is a menace to society,” he says. Still, the heart of his message is caution. “Chatbots may handle routine questions, but they cannot get inside my brain—or understand my patient’s real needs.” No app or algorithm, he insists, can read the whole story behind a symptom. 

Instead, Ganapathy believes that good care depends on context, conversation, and trust. “The human-trained brain understands not just symptoms, but a patient’s story—their social status, their context, and can factor in what matters most.” He’s wary of putting too much faith in technology and sees doctors as the guardians of real judgment. “AI can recommend, but only humans should decide management for real people, at real moments.” 

He calls for proper training, careful oversight, and honest conversations between doctors and their patients about what technology can—and cannot—be trusted to do. “A fool with a tool is still a fool,” Ganapathy says with a smile. “Technology is only useful in the right hands.” 

This episode reminds us that new tools are exciting, but real care is personal. As healthcare moves forward, it’s the human touch—and the wisdom behind it—that will always matter most. 

In line with his vision for advancing digital health, Professor Ganapathy is playing a pivotal role as Scientific Advisor for the upcoming Transforming Healthcare with IT (THIT 2026), South Asia’s leading international conference on telemedicine and digital health. Scheduled for January 30–31, 2026 in Hyderabad, India, THIT brings together global experts, innovators, and policymakers for keynotes, workshops, and interactive sessions designed to translate talk into real-world technology adoption. While Ganapathy is a staunch advocate of telemedicine, he emphasizes the importance of physical, face-to-face conferences in driving collaboration and meaningful change. His tireless efforts not only elevate the conversation but help bridge the gap between concept and impact, ensuring technology serves the cause of accessible, patient-centered care. For more details or to participate, visit www.transformhealth-it.org. 

The podcast is sponsored by DevelopmentAid. Procurement notices, funding and grants to opportunities, lists of potential partners, insights into market trends, databases of development professionals, webinars, latest news, and much more. Stay informed and connected.

Subscribe and Stay Connected

Transcript

Season 3. Episode 6: Beyond the Chatbot: Why AI in Healthcare Still Needs the Human Touch. Insights with Prof. Krishnan Ganapathy  

This podcast is sponsored by developmentaid.org, your gateway to international tenders, grants, and the largest job database in the aid sector. Secure free registration and access funding opportunities. Calls for partners, exclusive insights, and weekly News Digest. Become a member today to achieve your sustainable development goals. 

Hisham Allam: Hello everyone and welcome to DevelopmentAid Dialogues. I'm your host, Hisham Allam. Today we are talking about how artificial intelligence or AI is changing the way healthcare is delivered through telemedicine and telehealth. These digital services are becoming more popular worldwide, helping people get medical advice and care remotely. 

We will explore what AI can do and what it cannot replace, and what the future might hold for these important healthcare tools. Today's guest is Professor Krishnan Ganapathy, neurosurgeon and distinguished professor at the Tamil Nadu, Dr. MGR Medical University. He is an emeritus professor at the National Academy of Medical Science and a former WHO digital health expert. He also served as the president of Telemedicine Society of India, and the founder director of Apollo Telemedicine Networking Foundation. Professor Ganapathy, thank you for joining us again on DevelopmentAid Dialogues. It's a pleasure to have you back as we explore these exciting and important questions facing healthcare today. 

Krishnan Ganapathy: Thank you, Hisham for having me in this podcast. Before we go into the specifics, I have a confession to make. I used AI in getting the raw data for some of the highly technical questions, which I presume you're going to ask me in the next few minutes. However, after summarizing the world literature in these areas, I am predominantly going to use my God-given NI, natural intelligence, native intelligence to give you my personal perspective, which could be perhaps biased. 

I belong to the 20th century, trained in the BC era, not before the Corona, but before computers. Reviewing my five decades of hands-on experience in healthcare in India as a neurosurgeon and a digital health evangelist. My views will be primarily India centric, but then India constitutes 17% of the world's population. 

There is no world without India. Probably IT specialists of Indian origin have contributed 25 to 30% to the spectacular growth and development of AI. Our views therefore are relevant. I may be a cynic, but why are all the AI companies spending so much effort, time, and dollars? We are talking of trillions, not billions in promoting AI in healthcare. The answer is very simple. They want to milk a cash cow. CEOs of Apple, Microsoft, Google, and Amazon. In fact, most of the Fortune 100 companies have all publicly reiterated time and again that healthcare is perhaps AI's most pressing application naturally. After all, healthcare is probably the largest component of the economy and certainly recession proof. 

The ROI the AI companies are talking about is not just achieving better health; it is running all the way to the bank. Dimension market research predicted that AI in telemedicine spending alone is going to grow at a 26% compounded annual growth rate and will probably surpass 156 billion with a B by 2033. 

The percentage of physicians using AI has increased from 38% just two years ago to 66% one year ago, according to the American Medical Association, subway released recently. However, health cannot, in my personal opinion, be reduced to a series of algorithms above as sophisticated as they are. Medicine is not mathematics. It is not black or white. It is always various shades of gray responses to the same malfunction, to the same drugs could be different from individual to individual. The very fact that the Western world is spending billions of dollars in developing personalized medicine confirms this. Genomic analysis will be an integral part of individual healthcare soon. 

We are now in a stage of transition. Perhaps Gen alpha and Gen Beta in an AI world will not miss an eye. However, insofar as healthcare is concerned, humanoid chat bots cannot empathize and sympathize with a patient. The trained human brain has the capability to understand what exactly the patient wants, his her socioeconomic status, and can give proper weightage to the various variables which have to be factored in when a doctor communicates and talks to a patient. 

Hisham Allam: Thank you Doctor. That is an exciting beginning. I'm expecting an impressive discussion. To start off, what are your views on the use of AI and telemedicine?  

Krishnan Ganapathy: AI can be used in several aspects of telemedicine. The word telemedicine, the prefix tele very clearly indicates that we are using technology enabled remote healthcare to provide the digitally wherever there is a digital health divide to make any medical care available to the remote areas. 

Now, AI can be integrated into various virtual healthcare services, which enables a doctor to remotely diagnose, treat, and monitor patients using data-driven tools. By leveraging, the various aspects of AI we are all familiar with, computer vision, machine learning, natural language processing, MIOT medical internet of things and so on. AI has the potential to improve service delivery. Now, again, these systems can help optimize provider workflows, minimize wait times, directing patients to the most suitable care pathways. However, I have a point to contend with. This may be okay in a huge setup. The Kaiser Permanente, for example, one of the world's largest organizations today uses telemedicine more than face-to-face consultation. 

Again, AI algorithms can analyze medical images, CT, R X-ray, ultrasound, et cetera, remotely, and this would help a physician who doesn't necessarily have a radiological expertise. Another area we are talking of AI is AI powered chat bots. I personally am not very comfortable with chat bots because I think in spite of the very sophisticated programs, they cannot get into my brain. They cannot understand what exactly I want. 

Chat bots are excellent for handling routine patient communications, answer medical questions, and more administrative show like facilitating appointments, scheduling, providing medication numbers, reminding them what to do, where to go, when to go, how to pay, et cetera, et cetera. 

And of course, AI can enhance data security because today there is a cybersecurity threat. 

Hisham Allam: Focusing on the first touch point, with patients. Could you explore the advantages and disadvantages of AI for initial symptom checks? 

Krishnan Ganapathy: Okay, now, symptom checks as, as I said, I may be biased. But the real world is not mathematics. Medicine is not mathematics. A patient may come to me with 8, 10, 12 symptoms. 

Now, unless I ask the patient leading questions, many of the relevant symptoms may not be volunteered by a patient that is number one. Two, take the triage. What does a doctor do normally when there are five or six symptoms, the doctor decides based on his past experience, on his training, on his knowledge, he takes out one or two symptoms and pays emphasis on that. 

In the real world, no patient comes to you with a single malfunction, I don't wanna call it disease. I would rather call it a concern for his health. Nothing prevents a diabetic from having hypertension and a brain tumor, nothing prevents a person who had a head injury two months ago from now developing migraine and so on and so forth. 

In other words, coexistence of multiple pathologies is something which I really think the human brain and experience trained human brain at this point of time is probably preferred to a symptom checkup. There are, at least to my knowledge, three or four symptom checkers commercially available. I think many of you, the people who are listening to this may be familiar with WebMD, Ada, and so on. 

Among the three as well literature, the Ada app is a user-friendly tool, and I understand it's available in 130 countries. I don't think it's abuse in India, we have some Indian versions, which are now just been introduced into the market. There are almost 6 million use cases of Ada which about 45,000 patients per day. 

But even this, the most sophisticated has an accuracy of 70%. In other words, 30%, which is a huge unacceptable number. If you depend only on AI, you run a very high risk of making an inaccurate diagnosis. So, to the WebMD, they have compared it with other institutions, and they say to 62% as good, et cetera. Now what do, what does a human being do when it comes with a symptom check as now, of course, before I go into that, the advantage of a symptom checker is that it is available 24 by seven. You don't have to fix an appointment. You don't look at the watch and you know, only for a couple of minutes. But then the disadvantages, I think, outweigh the so-called advantages. The advantages which have been promoted by the symptom checker companies include faster triage, personalized insights, cost savings. 

Now this may not be relevant to a public health setting, particularly in a country like India, in a Western in the United States, many of my classmates are there. I understand a super specialist Teleconsultation may cost anything from 500 to 750 US dollars for 15 minutes of a super specialist time. 

So obviously you need to go to the right specialist rather than have one specialist say, okay, this is not exactly my specialty. You have to go to somebody else, and so on. So, a chat bot, a symptom tracker in that way will possibly, uh, lead you to the right specialist. But here again, the error in judgment could be as high as 30%. 

The most important disadvantages are the risk of misdiagnosis of technology barriers. Now, a major assumption which everybody's making is to have access to a symptom checker. You are digitally literate. You have power. There are no breakdowns of current electricity and so on. You are not only digitally literate; you are fairly familiar with how to use and navigate a symptom checker. 

This presupposes a reasonable idea of digital literacy, which not always available. Most important, I don't think a symptom checker will for the next several years at least handle complex or atypical cases. I always say that my patients do not read textbooks. What you, even the studies which you do in a controlled number on men clinic, statistically valid, excellent results you may get, but this was in a controlled end on men. Just doesn't work in the real world. 

Hisham Allam: But, professor, excuse me, we also have the human factor, the human misdiagnose for some cases, even because of the lack of expertise for , fresh graduate practitioners, right? 

Krishnan Ganapathy: Absolutely. 

Hisham Allam: So it is happening all the time with AI or without AI, we can have this kind of misdiagnosed for some cases. 

Krishnan Ganapathy: Absolutely. If I can diverge a little though, we are diverging when you talk of mental health. There have been several reports in the media. Not one, not two, at least a dozen, which I am aware of about suicides having actually taken place because of the inadvertent.  I'm assuming it's inadvertent of, I don't want to name any particular company. 

There is at least two companies against which a suit has been filed by the family, stating that when a discussion of mental health problems using a symptom checker, it ultimately resulted in the AI inadvertently, no doubt giving suggestions, which actually resulted in teenagers taking their own lives. 

So I think we are a long, long, long way, particularly when mental health issues are concerned. Of course, in, in all fairness, a symptom checker if, if the right questions are asked, and more important, the right answers are given. I'll give you a simple example. Tuberculosis is prevalent in India. 

Much more than many other countries in the world. Cough and no, nothing prevents a patient with active tuberculosis from also having a bronchogenic carcinoma. Again, I asked a clinician, I would ask the patient whether he was working in a mine. The patient may not think it is important enough to offer that. 

We have silicosis, we have minor lung disease, and dozens and dozens of other unusual things, but unless the right question is asked, the patient will not be able to give you the correct answer. So yes, the, the point I would like to emphasize is symptom checker is a great tool and we need to work on it, but we have a long, long way to go. 

The other thing is, one thing which worries me tremendously is the how did this AI validation come in the first place? I'm again talking of India to the best of my knowledge. Most AI, which is now being used in the Indian healthcare system, have all been validated in Western countries, meaning that the preliminary data was taken from Caucasians and there could be an inadvertent gender bias or discrimination the low economic group. In fact, there are several papers published on how, uh, African Americans have not been included in major studies and et cetera. So I think all these are important. The another important point is these AI algorithms were made maybe three years ago, maybe five years ago. Now, take India again as an example, in another five years TB will no longer be the major public health problem, which it is in India today. Today, if a patient comes to me from an endemic area, the AI will certainly flag a red flag, and the AI will say, this guy comes from the most populated, uh, district in India, where TB is very rampant, and therefore he would make a diagnosis of TB. However the same patient comes to me from the same district three years later when the incidents and prevalence of TB have significantly come down, TB may not be the correct diagnosis. In other words, your AI validation, the basis with which you do all this, have to be looked at again and again every couple of years. 

I am not sure this is going to cost huge amounts of money. I am not sure whether AI programs will be updated. In such large numbers every couple of years. 

Hisham Allam: Okay. That leads me my coming question, the same impact to the bigger picture. How is AI changing the way healthcare is delivered through telemedicine? 

Krishnan Ganapathy: Now, again, as I said to my knowledge. AI has made a tremendous influence in healthcare, but I am not really aware of AI making a major dent in telemedicine because see, it's, it's a paradox. I've always believed that technology is a means to achieve an end and not an end by itself. The role of AI in telemedicine, at least in the emerging economies is for a human being to appear on the screen and the patient hundreds of thousands of miles away. Most patients, including the educated ones, would still like a face-to-face consultation. 

I myself given a choice, I would like to hear from the cardiac surgeon who's telling me that I need a bypass stenting. I would like him to hold my hands. I would like to talk to him, not see him on a big screen. So when this is so, the role of AI in telemedicine will strictly be confined, which is what is also being done in India now for administrative purposes, for referring the patient to the right for, maybe, helping analyzing images, see diabetic retinopathy. We are already doing this in India in a big way. We have major projects with Microsoft and Google for reading chest x-rays, CT scans, et cetera. 

But the role of AI is only as a support, mainly an administrative support documentation recording. I understand that AI is CRI today, which are being regularly used in the Western world. Again, security,ensuring that whatever the transaction, which is occurring between the doctor and the patient is recorded in a hospital information system, in an electronic medical record, and so on. 

But I do not think, certainly not in my lifetime, AI will ever be used for a simple reason. There's no market. If I am a patient, already, I'm in trouble because the doctor is 500 miles away. And if I'm now going to talk to a human eye of 500 miles away, I will be very, very uncomfortable. 

So I don't think AI is going to be used in telemedicine at all except for many for a long, long time excepting for administrative purposes. 

Hisham Allam: You have a very critical point of view regarding AI, and you have, confessed this, uh, clearly in the beginning, and I totally respect this. But allow me to ask operationally, can AI help physicians manage high patient volumes in telemedicine? 

Krishnan Ganapathy: You are making an tremendous assumption here that patients have volumes and volumes and volumes to manage. Sir, in the real world, this is not true. There are thousands of doctors who are underemployed. There are thousands of doctors who are unemployed. If I were the health minister of the country or why were the bureaucrat in charge, I will spend more time in ensuring that the exist doctors are properly distributed. I'll give you a simple example as a neurosurgeon. This was a paper I published nine years ago. The numbers have changed, but it is essentially the same. 80% of India's neurologists and neurosurgeons do less work than what they're capable of doing because they all live in the metros, Taiwan cities, state capitals, et cetera. 

900 million Indians out total country population of 1.4 billion, even today live in areas where physically there is not a single neurologist or a neurosurgeon, so the emphasis is not in creating huge volumes. It just doesn't work like that and it, and it's not like buying a pizza. A patient wants to go to a particular physician because of his reputation, et cetera. He is prepared to wait for two months to get an appointment. He doesn't want a neural cardiac consultation. He wants to meet Dr. X, Y, and Z. So I think maybe this is relevant in many parts of the world, but I do know my own classmates, my own friends, my own contacts. I really don't think there are huge volumes, as is generally believed. 

Yes, there are waiting volumes, but the volumes is because of this cash city of doctors. Not because an individual doctor. So I would spend all my effort and time and technology in ensuring that underemployed doctors, unemployed doctors are users, number one. Number two. 

Hisham Allam: Sorry, doctor , allow me to interrupt here. But there is a lot of countries according to WHO that are suffering from lack of doctors like South Sudan, Sierra, Leon, and the United States, even South Korea, and most of Europe countries, are facing, lack of doctors. So this is somehow something, worldwide, even if, uh, India is exactly exception. But, it's happening. 

Krishnan Ganapathy: No, no, no, no, no. Sorry. Maybe I said the wrong thing. You hit the nail on the head. So we use technology not in increasing volumes for doctors. In my opinion, should not see more than six, seven patients and other, so when I say increasing volumes, what I mean is trans border. 

So we should be spending our time in making the underemployed doctors make them available in Africa, Sierra, all the countries you mentioned 

What I understand now is the, the present AI is being used to make a single doctor to see more and more patients. I don't agree with that. I think a doctor should not, see more the next number of patients for the simple reason he needs to, establish a contact with the patient. 

When I talk of volumes, I'm talking of a single doctor, managing more patients in his limited time using AI, which is being done, but I'm not sure that this is the right approach. That's what I meant. 

Hisham Allam: Okay. This is clear now. Looking upstream to prevention. How does AI promote preventive health through telemedicine? 

Krishnan Ganapathy: Okay. I think this is a great area and I think preventive health through telemedicine can be used in large ways in a very simple way. I used NI by promoting health literacy about 15 years ago, but I had to stop it after four, five years. So AI can be used in promoting health literacy, patient empowerment, knowledge empowerment. 

NHS has got a major system on this, and in fact, the chief knowledge officer of NHS 20 years ago made a great statement. He said, well, pay promoting health literacy is more powerful than antibiotics or surgery in healthcare. I totally agree. But AI can help translate languages. For example, let's say I have 30, 40 modules dealing with diabetes, hypertension, that thing, this thing, this thing, et cetera. 

AI can be used in converting it to regional languages, and AI can be used to deliver reliable authenticated health information across the length and breadth of many, many countries, which a doctor would not be able to do single-handed, of course, a doctor assisted by AI. This will make a major difference. So this is one.  

Two, early risk detection, predictive analytics. Predictive analytics are now become extremely careful and telemedicine can do this in a great way. So, for example, the human being may not be able to analyze my last five years of blood sugar. But AI can predict with a very high accuracy that I'm likely to develop complications of A, B, C, et cetera. So, all this could be deployed in huge numbers in rural areas, suburban areas, et cetera. And this final information, given to a human being. Real-time remote monitoring. This is another area where AI, I was surprised to find that India ranks number three in the whole world in accumulating data from variables. 

I don't know how reliable this information is, but I was absolutely surprised. So variables I do know, have become very common even in developing countries like India. So we are bombarded with data. So, and this variables obviously are won by a patient who is several hundred thousand miles away from the doctor apart of telemedicine. 

So real time remote monitoring of voluminous amounts of data can be done using AI conclusions drawn and the conclusion alone, uh, sent through, uh, your internet network to the doctor. So this is something which is very helpful. Then of course we are talking a population based preventive care, public health. 

See, AI in public health can play a very, very major role, and the word telemedicine is used in a loose sense of the word, where a public health expert physically located in one area using AI can institute preventive measures in regions far, far away from that . 

Hisham Allam: When AI suggestions a conflict with the clinical judgment, what escalations or override protocols should be in place? 

Krishnan Ganapathy: Okay. There are two words here. One is protocol, another is regulations and legal aspects. Now all countries are starting this. India has just started this, but unless there is a test case in the court, I think the rules will still be a dark area still. 

But at the end of the day, the gut feeling, I mean, the Harvard Business Review published a paper saying that CEOs of Fortune 50 mergers and acquisitions ultimately depend on gut feeling, not on the volume statistical data. Similarly, as a clinician, if I have a gut, it's difficult to describe this, but if I think I'm not very comfortable with your opinion, AI given, I think we should go deep into it. So there should be a protocol that the a AI is at the best can assist a clinician. It'll make me think of things which I normally may not have thought off. 

Absolutely right. But I am very, very uncomfortable and I think most clinicians are uncomfortable in, in an AI can read an image. Far better than me and AI can look at a pathology slide and make out changes, which a human being cannot make out. But whether I'm going to use that, whether I'm going to decide on what I'm going to do from a management perspective, the AI can recommend management protocol. 

But the ultimate decision on whether this management is to be implemented for this particular patient at this point of time, it should be entirely left to clinical judgment because the AI programs, to the best of my knowledge do not take into account what the patient wants. I'm eight to five years old. I've had an excellent quality of life. I've been diagnosed with multiple secondaries. I do not want any more treatment. All the treatment recommended by AI is proton therapy, which in India costs 50 of rupees and the, the insurance company does not bear this. I think it's a crime. I would not even tell my patient that you have such an OB treatment option available because I know that this patient will never be able to raise that funds. So this is where clinical judgment comes. I tailor make personalized customize what AI tells me to suit that particular individual. And, uh, so I think the protocol should make it very clear that ultimately it is a clinician who uses AI , but the final judgment is based on a discussion with the patient by a human being and not by the world's most intelligent. 

Hisham Allam: Turning to the guardrails and risks you have raised about chatbots. What are the key limitations of AI chatbots in healthcare? 

Krishnan Ganapathy: First of all, the comfort. There are not many papers of how comfortable I am in dealing with the chatbot. 

As I said, chatbot is great for handling patient communication, facilitating appointment during, and things to, uh, provide medication reminders. This is a great area where a chatbot can be of tremendous help, but again the best chatbots in the world do have limited medical expertise and diagnostic accuracy. 

And finally, the most important, the lack of human empathy and contextual understanding the context in which it is done. The same problem, the same symptom in Chicago is totally different from what it is in a small town in India or in, uh, in a very, very primitive place in Africa. So the context chatbots can never ever replicate. 

The emotional support, which I am able to give trust building, see a patient and a doctor in my days in my generation, was a question of trust. And unless there's trust building, there are enough scientific papers to show that when there is trust there, results are different, become better, even when the same medicine is given the same radiotherapy and chemotherapy is given. 

There are a lot of things which we still don't know, and again. Patient safety concerns. I don't think a chatbot will be able to ask the right questions to the patient, which a doctor can give. And what worries me is the multinational companies are promoting chatbots so aggressively and extensively. 

There's a risk of over reliance. In India we see, boards in the big roads signed boards saying that this particular corporate hospital now has an AI, uh, cardiac evaluation system, master health checkups,AI, AI everywhere. They use AI. In fact, uh, just two days ago I read in the newspaper that all the big corporate hospitals in India have invested almost 20% of their budget for the next 3 years on AI facilitated things. They, everybody wants AI and you know, this is a very much concern to me, and unless we have more data, we are in a stage of transition. Data, privacy, security, all these problems will increase when you start using more AI. 

And of course, finally, ethical and legal complications. I don't think any country in the world has clear cut directions. Now, if I use AI and something goes wrong with my patient, patient files a suit, can I blame the AI? Which developer the AI is going to be involved as a correspondent? 

Alternatively, if I don't use AI and something goes wrong, can the patient sue me? For not using AI. So these are all problems which are not going to be answered unless there are test cases. So I think it'll be a long, long time before we actually, uh, have clear cut answers. 

Hisham Allam: Professor Ganapathy, finally, and briefly for our audience, what the take home message you would like to send? 

Krishnan Ganapathy: Well, that's a real difficult question, but then a clinician who's not AI literate it's a menace to society. I honestly think it's a menace to society. If I was the authority, I would ensure it is happening, but I would put this on fast track more. I would ensure that every healthcare provider, be he the vice Chancellor, or be he or Ward Boy is taught at his level on the use of AI. AI enabled clinicians will be far better than a healthcare provider who is naive about AI, uh, healthcare delivery system, which does not understand and accept the limitations of deploying AI is also doomed for failure. Trust is a key word in healthcare. A combination of AI with various other factors like wearables, IOT devices, telemedicine platforms, et cetera, is going to make a tremendous change. Regulatory approvals, infrastructure gaps all this have to be addressed. 

I would like to conclude by being a little philosophical, a quote from last Lexel, the inventor of the gamma knife, which at that time was the world's most sophisticated medical technology. He said, I quote, A fool with a tool is still a fool. I belong to a generation where we were taught that technology is a means to an end and not an end by itself. 

I shudder to think of the day when the good old fashioned 20th century doctor, whose primary reason for existence was to provide tender loving care and wipe the tears of the patient, becomes an endangered species. Maybe AI enabled humanoids will have algorithms where the algorithms which incorporate empathy, sympathy, and virtually wipe the tears off. 

And I'm very concerned that the aggressive marketing of AI will I hope, will never, ever replace God given eye. 

Hisham Allam: Thank you Professor Ganapathy for sharing your invaluable expertise on the evolving interplay between AI, telemedicine, and telehealth. Your insights illuminate both the possibilities and responsibilities as we navigate this new frontier of healthcare. 

To our listeners, if you found this conversation insightful, please follow DevelopmentAid Dialogues for more discussions that matter, and share this episode with your network. This has been your host Hisham Allam. Until next time stay, well and stay informed. Goodbye!