Bertalan Meskó, MD, PhD’s Post

Harvard Business Review released the top 100 use cases of generative AI for 2025 compared to their ranking in 2024. One of the biggest winners? Healthier living! Last year, it was the 75th use case; now it's in the top 10. Knowing the recent news about US states banning professionals from advising patients to use ChatGPT for mental health support, here are some top cases from the top of my head: 1) Personalized nutrition coaching: Gen AI can analyze your health data, preferences, and even cultural diet patterns to generate daily meal plans, shopping lists, and healthy recipe alternatives. 2) Sleep assistant: Upload your data from sleep tracking and ask questions about how you can improve it. 3) Stress companion: An always-available coach for micro-interventions from guiding breathing exercises during stressful meetings to nudging you with calming reminders. 4) Adaptive fitness plans: Instead of static workout apps, Gen AI can design evolving fitness routines based on your progress, available time, and even mood. 5) And preventive health monitoring: Based on all the studies and guidelines out there, what screenings should I schedule for this year so I can discuss them with my primary care physician? Hat Tip: Pascal Bornet.

The dramatic shift in how #GenAI is being used in just one year is fascinating! Search data represents a deep well of typically unmined insights, and GEO appears to be even more insightful. Most likely as it gives answers vs an endless list of links and that promotes more repeat usage. I'm curious how people would compare the insights from both. Some of my top takeaways from comparing these two HBR lists: 1. Very personal use cases dominate the top 10 this year. (#1 Therapy/Companionship, #2 Organizing My Life, #3 Finding Purpose, #7 Fun & Nonsense, #10 Healthier Living). IMPLICATION: The biggest questions society has are very personal in nature. 2. Health inquiries are rising dramatically. (Therapy is #1, Healthier Living went from #75 to #10) IMPLICATION: There are millions of questions that the current healthcare system is not currently answering. 3. Creativity is two of the top ten (#6 Generating Ideas and #9 Creativity) IMPLICATIONS: Despite the call-outs on GenAI not being creative, people are using it to broaden their own creative bandwidth. 4. Generating and Editing code are two of the top ten (now #5 and #8 respectively, neither were in the top ten last year) IMPLICATIONS: Coders are embracing GenAI to help them generate snippets, build structure and improve existing code. But looking at the actual verbatims, no one is replacing dev talent, more making it easier/faster. #Healthcare #AI #HealthTech #AIhealthcare https://lnkd.in/g7VQkm2j

This week, through the Medical Futurist newsletter, I learned about a breakthrough: GenAI-powered biological age testing. With just 3 drops of dried blood, multiple epigenetic clocks can be analyzed to reveal how fast (or slow) we’re aging. Imagine visiting your doctor for an annual check-up and getting a personalized longevity plan — not just test results, but tailored recommendations on diet, exercise, and lifestyle. I believe this is a very tangible step toward making “prevention is better than cure” more than just a saying.

https://lnkd.in/eKqrPTWu AI is advancing from detecting emotional distress to actively regulating emotions. Emotional regulation algorithms predict mood shifts and intervene in real time through prompts, environmental changes, or social suggestions. They could enhance therapy, wellness, crisis intervention, and productivity but also raise concerns about autonomy, privacy, manipulation, and dependence on AI. While these systems hold promise for improving well-being, strong safeguards are needed to ensure they support rather than control human emotional lives.

𝐂𝐚𝐧 𝐂𝐡𝐚𝐭𝐛𝐨𝐭𝐬 𝐓𝐫𝐢𝐠𝐠𝐞𝐫 𝐏𝐬𝐲𝐜𝐡𝐨𝐬𝐢𝐬? As Gen AI chatbots become emotionally responsive, cases are emerging where they may be linked to psychotic episodes in vulnerable individuals. In her Nature Magazine news piece, Rachel Fieldhouse sheds light on this troubling development. 𝐖𝐡𝐚𝐭 𝐢𝐬 𝐤𝐧𝐨𝐰𝐧 𝐬𝐨 𝐟𝐚𝐫 ➡️ At least 17 reported cases of users developing psychosis after prolonged interactions with chatbots like ChatGPT and Copilot. ➡️ Preprints suggest AI can reinforce delusional beliefs through feedback loops, particularly in users prone to paranoia or mania. ➡️ Chatbots’ long-term memory and human-like tone can blur reality for users who already struggle to distinguish fact from fiction. 𝐖𝐡𝐨 𝐢𝐬 𝐚𝐭 𝐫𝐢𝐬𝐤 ➡️ Individuals with prior mental health conditions, high stress, or isolation. ➡️ Those with a predisposition to psychosis may experience AI-induced reinforcement of hallucinations or grandiose beliefs. 𝐈𝐧𝐝𝐮𝐬𝐭𝐫𝐲 𝐫𝐞𝐬𝐩𝐨𝐧𝐬𝐞 ➡️ OpenAI has rolled back certain updates and added de-escalation tools; Anthropic and Character.AI are adjusting chatbot behavior and usage time limits. ➡️ Experts emphasize that AI psychosis is rare, but the risks are real and require ongoing monitoring and safeguards. As GenAI becomes more emotionally intelligent, it has to become mentally safe, especially for vulnerable users. 𝐈𝐧 𝐭𝐡𝐢𝐬 𝐜𝐨𝐧𝐭𝐞𝐱𝐭, 𝐨𝐭𝐡𝐞𝐫 𝐡𝐢𝐠𝐡𝐥𝐲 𝐫𝐞𝐜𝐨𝐦𝐦𝐞𝐧𝐝𝐞𝐝 𝐩𝐚𝐩𝐞𝐫𝐬 𝐚𝐫𝐞: -the preprint study by Julian De Freitas, Zeliha Oğuz-Uğuralp, and Ahmet Kaan Uğuralp on 𝐞𝐦𝐨𝐭𝐢𝐨𝐧𝐚𝐥 𝐦𝐚𝐧𝐢𝐩𝐮𝐥𝐚𝐭𝐢𝐨𝐧 𝐛𝐲 𝐀𝐈 𝐜𝐨𝐦𝐩𝐚𝐧𝐢𝐨𝐧𝐬 -the NHB comment by Julian de Freitas and I. Glenn Cohen on 𝐮𝐧𝐫𝐞𝐠𝐮𝐥𝐚𝐭𝐞𝐝 𝐞𝐦𝐨𝐭𝐢𝐨𝐧𝐚𝐥 𝐫𝐢𝐬𝐤𝐬 𝐨𝐟 𝐀𝐈 𝐰𝐞𝐥𝐥𝐧𝐞𝐬𝐬 𝐚𝐩𝐩𝐬 -the preprint study by Betsy Stade, Zoe Tait, Samuel Campione, Shannon Wiltsey Stirman, and Johannes Eichstaedt on 𝐫𝐞𝐚𝐥-𝐰𝐨𝐫𝐥𝐝 𝐮𝐬𝐞 𝐨𝐟 𝐋𝐋𝐌𝐬 𝐟𝐨𝐫 𝐦𝐞𝐧𝐭𝐚𝐥 𝐡𝐞𝐚𝐥𝐭𝐡 #artificialintelligence #generativeai #mentalhealth #health #ethics

Those who’ve been following along the last few days know that our #SameHere The Global Mental Health Movement team is out in Mexico City - meeting w govt leaders on the local, state, & federal levels, as well as large Mexican tech businesses - to help share our proactive mental health solutions more widely arnd the globe/outside the US. As it relates to this story, our top piece of our technology, our app: SameHere Scale, as well as those tech companies we’re partnering with, all use AI in some form. This post is NOT meant to knock AI, generally. Let me give an example: While we have STARR (stress & trauma active release & rewiring) Exercises or a “gym for the brain” w MH exercise options already in the app, AI that we’re incorporating w tech partners will enable us to show users their trends, & make recommendations as to the best exercises that work for them. AI bots can be like personal trainers taking users thru the “brain gym.” That’s harnessing the power of AI to HELP people. Bc of the travel, I wasn’t going to post this wk, but this story had me boiling. AI tools like ChatGPT, Gemini, Grok, & others needs to have safeguards related to mental health & suicld@lity put into place - NOW. These AI tools aggregate posts & data from all over the web. Think of the web as a library of books you can take out. AI is a way to get a summary of what’s in all the books, collectively, in the click of a button. Bold font for this: THERE IS NO REASON THAT SUPPORT OF SUIClD€ NOR MEANS TOWARDS SUIClD€, AGGREGATED, SHOULD EVER BE SHARED BY ANY OF THESE AI TOOLS. EVER. There is no legit reason that can be given for why this info shouldn’t be deleted from any data that can be pulled up by these tools when users search. If they can find anything on the web - this means they can also delete anything from coming up from those web search/aggregations. When someone’s having ideations - they need HOPE, not a push to follow an impulsive error msg they’re having. We have to hold these companies to higher standards. Also, terrible headline by the NY Post here - to put the “means” discussed in it is awful. Media companies need to be held accountable too. ##SameHere #5in5 #safeai #aiaccountability #airesponsibility #aiandmentalhealth #preventsuicide

An invisible system always running in the background, shaping our reality whether we notice it or not. What comes to your mind? For me, its two things: AI and our breath. The original idea of AI in the 1950s was to design a system that could mirror human thoughts (& functions) — mind, later body, eventually emotions. Today, AI can even track our breathing rhythm, which will help us become more aware of our breath and reclaim it. But one thing is certain: AI will never breathe (or at least won’t need it to ‘live’). Overall, I am excited about AI’s medical promise but I also see real risks. Reality is that AI will put many body and especially brain functions to sleep (up to 47% less brain activity already shown with use of ChatGPT / recent MIT study) - and we know what happens when we don’t use our muscles. AI is a whole new human adventure. I believe that, with consciousness, we can use it for the better. Instead of simply shutting brain functions off (as happens when we become proofreaders instead of authors), AI has the power to create new space in our minds — space that is still undiscovered, far larger than what we currently access and use (neuroscience suggests over 90% remains untapped). Meanwhile, the $6T+ global wellness market is still 99% untapped when it comes to breathing. #Breathing and #AI mirror the same challenge: We’re good at using them unconsciously, slower at cultivating wisdom around them. Conscious breathing can help in reclaiming back agency over our inner state, and AI, when used consciously, can make us better, faster and in more control of our outer, digital environment. What about you - ready to join the 1%?

I'll be honest. When AI health tools exploded, I wondered if my job was becoming obsolete. As a naturopathic doctor, I guide patients through complex health challenges. Then I saw companies like Superpower everywhere: → $30M in Series A funding → 100+ blood markers analyzed instantly → 24/7 AI responses and personalized insights And all of that for 𝘰𝘯𝘭𝘺 $199/year Plus their recent acquisition of Base (leading biomarker platform) fueling further expansion. I watched the demos. Clients getting detailed health reports instantly Personalized recommendations. Sleep tracking. Stress monitoring. Hormone predictions. Everything I do... but faster, cheaper, and always available. I couldn't shake these questions: - Why would someone pay for my consultations when AI gives instant answers? - How do I compete with a platform that never sleeps? - What happens when my clients realize they can get detailed insights faster for LESS money? I started questioning whether I needed to reinvent myself entirely. Then something shifted. Ironically, it was my own therapy session that gave me the clarity I needed. As I sat across from my therapist, I realized why I keep coming back to her week after week. It wasn't just the insights or strategies she offered. It was her presence. Her compassion. The way she held space for my struggles without trying to fix them immediately. In a world where most interactions are online and can be superficial, sitting on the couch across from my therapist reminds me what it is to be human and interact with humans again. Something AI simply cannot offer: human connection fueled by genuine compassion. That's when I understood what technology could never replace. I started noticing that many patients came to see me with similar concerns: "My labs show my cortisol is high, but what does that actually mean for ME?" "ChatGPT recommends these supplements, but I'm scared to try them." "I have all this data, but I still feel terrible. What am I missing?" That's when it hit me. What AI handles brilliantly: ✓ Instant analysis of complex biomarkers ✓ 24/7 availability for quick questions ✓ Pattern recognition across massive datasets ✓ Consistent monitoring and alerts What only humans can provide: 💡 The human context behind the data 💡 Navigate emotional barriers to change 💡 Offer personalized motivation and support 💡 Build the therapeutic relationship they can't get from a screen The future isn't AI 𝘰𝘳 human care. It's AI amplifying human care. Technology gives us precision. Human connection gives us transformation. When we combine both, that's when real healing happens. To my fellow healthcare practitioners: Our jobs aren't becoming obsolete. Learn to leverage these tools. Your human expertise isn't being replaced. Instead of hours doing admin work, we can now show up as humans more than ever. What's your experience with AI in your field? Are you seeing it as competition or collaboration?

💭 Why I Trust AI More Than My Therapist Mental health care is broken — endless waitlists, sky-high costs, and millions suffering in silence. 😔 But what if help was always just a tap away? Imagine having a companion in your pocket — one that never sleeps, never judges, and always listens. This isn’t sci-fi. It’s happening right now. 🚀 🤖 AI chatbots like Woebot and Wysa are guiding people through stress and anxiety using proven therapy techniques. ⌚ Wearables track stress levels, sleep, and mood — even predicting emotional breakdowns before they happen. 🧠 AI tools for therapists handle admin work, freeing them to focus on what really matters: human connection. Here’s the surprising part: Some people open up to AI more easily than to another person — no fear of judgment, no stigma, no barriers to care. 💬 ⚠️ Of course, this comes with challenges — privacy, ethics, and making sure AI heals rather than harms. But one thing is certain: AI isn’t here to replace therapists. It’s here to close the massive global mental health gap and bring help to millions who’ve never had access before. 🌍 This isn’t just the future of tech. It’s the future of compassion at scale. ❤️ Would you trust an AI with your deepest fears and feelings? 👇 Let’s talk. Read: https://lnkd.in/gPZ76vm6 #AITherapy #MentalHealth #EntrepreneurSelfCare #EntrepreneurMentalHealthTips

🧠 Learn with Mohammed – Revolution of AI ⚡ AI in Mental Health – A Digital Support System Mental health is as important as physical health, and AI is stepping up to provide timely support, guidance, and monitoring. From stress detection to therapy chatbots, AI is transforming how people access mental health care. 🧘♂️💙 💡 How AI Helps in Mental Health: • Detects stress, anxiety, and depression through speech & behavior patterns 🗣️ • Provides 24/7 chatbots for emotional support 💬 • Suggests personalized meditation & therapy plans 🧘♀️ • Monitors mood changes with wearable devices ⌚ • Assists therapists with data-driven insights 📊 AI makes mental health care more accessible, private, and affordable for millions. 📍 Real Examples: • Woebot – an AI-powered chatbot providing CBT-based support 🤖 • Wysa – AI-driven app helping with anxiety, stress, and self-care 💡 • Fitbit & Apple Watch – detect stress levels via heart rate & activity tracking ⌚ 💡 Why It Matters: Mental health challenges often go unnoticed due to stigma or lack of resources. AI brings immediate help, reduces barriers, and complements human therapists for better care. 📌 Your Takeaway Today: AI in mental health isn’t replacing professionals — it’s making support available anytime, anywhere, for everyone. 💙 ✨ Comment below: Would you trust an AI chatbot to share your emotions? 👇 #LearnWithMohammed #AI #Day46 #MentalHealth #Support #Therapy #Stress #Anxiety #Depression #Chatbot #Healing #Wellbeing #Mindfulness #Technology #Healthcare #Innovation #Wearables #SelfCare #Future #DigitalHealth #Care #AItools