Jan Beránek’s Post

Apps have been created help identify if an AI most likely wrote a written text, but their results are not always correct. There are also indicators that you, as a human reader, can look for on your own. For example; - The introduction and the conclusion are usually my first clues. They are a perfect restatement of the issue (too perfect) and often sound quite robotic. - Sometimes the author forgot to remove words such as "Certainly!", which ChatGPT often writes at the beginning of its responses. - People have also noticed repetitive words from text written by ChatGPT and other LLMs. Dr. Nguyen did a study where he looked at the word "delve" in medical studies. There is a definite increase over the past two years. Does that mean that every study with the word "delve" was written by ChatGPT? No, but it's worth keeping in mind. Should medical writers stop using the word "delve"? No, ChatGPT learned it from content created and posted by humans. It's acceptable to use the word "delve". #ai #artificialintelligence #Chatgpt #llm #llms #medicalwriting #medicalwriter #healthcomms #healthcommunications #pharmamktg #pharmamarketing #writing #genai #generativeai #medical #science

Are medical studies being written with ChatGPT? Well, we all know ChatGPT overuses the word "delve". Look below at how often the word 'delve' is used in papers on PubMed (2023 was the first full year of ChatGPT).

Are medical studies being written with ChatGPT? Well, we all know ChatGPT overuses the word "delve". Look below at how often the word 'delve' is used in papers on PubMed (2023 was the first full year of ChatGPT).

Every Monday I used to get a csv report from our phone systems showing call times, pickup rates etc, but many of our team members have multiple extensions so collating the info took about 20 mins per week. And that was not an exciting 20 minutes to start the week with! Now, more than ever, we’re trying to streamline everything we do with AI; partly to acclimatise to the new tech landscape and to save ourselves some headaches too. I’ve now built a custom GPT inside ChatGPT4. Drop the old spreadsheet in and the nicely formatted data comes out in 20 seconds or less. Time saved: 19m 40s each week

AI was mentioned by President Biden in the state of the union. The Internet was first mentioned in the state of the union by President Clinton on January 27, 2000. What does this mean with respect to how fast things are progressing? For the economy, stock markets and adoption? Does it just mean that we are more aware of the impact of software on the economy and hence it’s mentioned earlier? Or does it mean that we are further along in the creation of a bubble than the timeline between ChatGPT release in late 2022 and the present day would indicate?

ChatGPT outperforms physicians in empathy, usefulness, correctness, and potential harm... If your doctor is not (yet) using ChatGPT as part of their routine analysis, would you still trust them? With evidence so strong, is it still ok to disregard when lives are at stake? What would need to be true for us to have much more constructive discussions with our physicians about our health - armed by the insights our AIs have about us? It seems like the future is already here... just unequally distributed... #PracticalAI #FutureOfWork #AIHealth

What would you do if you had an extra week each month? With our ChatGPT course, we provide you with the tools to not only understand and effectively use artificial intelligence but also to transform the way you work. Imagine having the ability to automate laborious tasks, freeing up time that was previously spent on manual processes. By harnessing technology, you can gain not just hours but even additional weeks per month. Register for our free course: https://lnkd.in/gMq4BYNR

Ever found yourself wishing for a magic wand to navigate the maze of regulatory compliance? 🌟 Imagine asking any regulatory question and receiving clear, precise answers within minutes. Well, wish no more! Introducing ChatMDR 2.0 – your AI-powered ally in turning complex compliance challenges into straightforward tasks. This February, we're not just imagining; we're delivering. Experience our Smarter option with 200 free questions and see how quick and insightful answers can transform your regulatory workflow. Why wait? Make your regulatory wishes come true today: https://lnkd.in/eGgwbUn6 Dive deeper into the future of compliance: chatMDR.eu #ChatMDR2 #RegulatoryCompliance #MedicalDevices #AIComplianceTool

I'm touched by the positive feedback on ChatMDR and the enthusiasm from all of you. ⭐️⭐️⭐️⭐️⭐️ Thank you to everyone who's been part of this journey. If you haven't yet, don't miss the chance to try ChatMDR 2.0 for free this February. 🔗 https://lnkd.in/eDnVmy_n #ChatMDR #RegulatoryCompliance #artificialintelligence #medicaldevices #medicaldeviceregulation #AIforMDR

In mentally preparing for the UACOM-P panel on medical misinformation impacting patients, I was thinking about my first experience with patient's using ChatGPT. I remarked "wow you have great questions!" She sheepishly said, I checked ChatGPT last night. I said, "That's great." So, what do we tell our patients that are using CHATGPT or other LLMs? 1. It's OK to use it. It can be satisfying and helpful because it's often correct, engaging & interactive, always available, and can be compassionate. 2. It works by word prediction, so it doesn't understand what it says 3. It acts like an overconfident friend; it can be persuasive despite being wrong and biased. And how do we protect ourselves? 1. Ask if it would be a problematic if it was wrong, and if so beware 2. Corroborate with experts and trusted people 3. Acknowledge A.I. bots using this tech, have been used as "weapons of mass disinformation" on the web and social media.