Marielle van de Pol

OpenAI's CEO Sam Altman Predicts AGI Could Appear Within Five Years, With Far Less Impact On Society Than Most Expect https://buff.ly/3AqC845 🚀 Artificial General Intelligence (AGI) Within Five Years? OpenAI CEO Sam Altman predicts AGI's arrival by 2029, sparking debates on its feasibility and implications. 🤖 AGI: The Next Frontier or Overhyped Dream? While Altman envisions AGI revolutionizing industries, skeptics question the timeline and potential risks. 💡 Innovation vs. Hype: Is the tech world on the brink of a breakthrough, or are we setting ourselves up for disappointment with ambitious AGI forecasts? ⚖️ Balancing Progress and Caution: As we chase AGI advancements, are we adequately addressing ethical concerns and potential societal impacts? 🌐 The Future of Work: With AGI on the horizon, how should professionals prepare for a landscape where machines might outperform human intelligence? So in summary, let us define what AGI first because i am still not clear on the definition. you can not reach something that is not well defined. I agree that we do not need any other breakthrough to reach great level of AGI but we need to know how we can harness what we have to make our systems more powerful. #AI, #GenAI, #GenerativeAI, #AGI

6

1 Kommentar

Are you familiar with Large Language Models, Entity Relationship Diagrams and training of neural network? This is really valuable for a data leader. But it is not sufficient to get the organisation on your side. You need sincere interest in people, you need empathy, you need creativity and openness. Not really data skills, right? I'd be happy to read your comments. My bi-weekly "Food for Thought" series invites you to share your ideas. #data #cdo #datamanagement #foodforthought

17

2 Kommentarer

A few days ago, a colleague of mine asked me how AI could help support patients better, and especially help improve medication adherence in a sustainable way. Addressing this issue is extremely complex as it requires to deeply understand each individual (personality, history, beliefs, and behaviors) as well as his/her circles in life, and other external factors and see how we can leverage all of this to build an approach that resonates and induce new behaviors. LLM can definitely offer new perspectives in that space. Have you ever tried to get advice from ChatGPT in complex life situations? Do it! You need to build a very rich prompt as you need to map the different stakeholders, their personality, their motivation in life, their connections, etc. In the professional world, you also need to take into consideration circles of power, informal relationships, etc. I used it with professional and personal situations. I wanted an external point of view on how to behave in specific complex situations. The results ChatGPT gave me were ABSOLUTELY AWESOME, unexpected, and even… disturbing. Of course there is a bias as it depends on a prompt that I built based on my own perceptions of situations. (Is that so different from asking a good friend for advice?). Back to the topic of medication adherence, I think that future LLM-powered digital solutions have the potential to provide very innovative approaches by truly taking into account what makes each of us unique. Such AI coach will be some kind of permanent companion, fueled over the years with all our interactions with the world as well as our emotions at each instant, and will be able to activate the right triggers at the right time. By understanding us as unique individuals, AI will outperform any doctor to drive new behaviors over the long term. The future of patient coaching will be definitely far from the simple ‘pill reminders’ we have today. 😉 Sharing an article on this topic that I found inspiring. https://lnkd.in/e9YgaY7r #adherence #medicationadherence Liana Lubel Oana Butu #acare #patientcentricity

33

1 Kommentar

It’s curious that we don’t hear much about AI explainability challenges anymore - perhaps the AI explainability "champions" have now become the champions of AI doomerism 😀 Explainability remains critically important for AI systems. There has been remarkable progress in the last decade, with much more on the horizon. However,here’s a friendly reminder for context: ✈️ We don’t fully understand why planes stay in the air 🤔 💊 For many drugs, we don’t know exactly how they work 🤔 Something to think about next time you board a flight or take a pill 😊 I’d bet we’ll achieve explainable AI long before solving these other mysteries. #AI #GenAI #Explainability #LifeSciences #Pharmaceutical #ClinicalTrials #Medicines #Flying #HealthTech #AIinPharma

6

3 Kommentarer

🤔Does your organization use centralized or decentralized AI governance? An interesting article from MIT Sloan explores how the effectiveness of a centralized approach depends on your organization’s AI maturity. 📊 At the very early and very mature stages, centralized governance may add limited value. However, during the experimentation, integration, and scaling stages, a centralized approach can have the biggest impact by aligning efforts, setting standards, and driving cohesive growth. 🚀 What stage is your organization in, and how are you approaching AI governance? 👇

5

The answer is: if you haven't thought about it earlier, you will increase your rate of bad decisions. 😱😓 The question... What is the importance of strategic thinking, strategy planning and strategy implementation? It was with that mindset that I joined earlier this year the Strategic Management and Innovation specialization by the Copenhagen Business School. Strategy has been part of my life... ↳ through school, including my two MBAs ↳ through work, given my management & leadership positions ↳ personally, as I always enjoyed thinking beyond the "what" I am doing and diving into the "why" and "how" I am doing it. It's enough to say... that it paid the highest dividends in my life so far 🤑💰💸 This is the ONE reason why I will recommend this course. If you are interested in the topic. Or just having some time to invest on yourself. If you want to... ↳ Find different approaches to life ↳ Understand the rationale behind business decisions ↳ Ensure that good thought is behind your business ↳ Learn how to move from thinking into execution ↳ Absorb a number of models and concepts This is for you. I was particularly impressed by the Strategy Management course, the first one, as it dared to challenge established (and old fashioned) models. And the capstone project. It was fun! 🤓 🛑 I was less impressed by the Strategy Formulation, because it much ignored all that was said in the Strategy Management and got back to the 80s models and all I have learnt 20 years ago. 🛑 Even less impressed by the Strategy Implementation, as it was a bit all over the place, which is proof in itself that moving from formulation into implementation is not at all easy! One thing to remember from the course? 💡That the best novel strategies are flexible, resilient, ever-adapting and developed my the entire company, with inputs from all levels. Top-down, large scale formulation by a team of elite consultant is a thing of the past and of companies aiming to fail. How do you feel about strategy?

54

45 Kommentarer

As we head into the weekend, remember: - Not every problem requires an AI-based solution. Don't bring a bazooka to a knife fight 🤔 - You don't need an AI Strategy - you need a Business Transformation Strategy powered by AI 🚀 #AI #BusinessTransformation #LifeSciences #GenAI #Innovation #Leadership #DataScience #lifesciences #pharmaceutical

16

2 Kommentarer

Great article on the value of real-world data, patient empowerment, and the general need for data quality improvement. The pharma industry can also contribute to a better understanding of the patient journey by focusing on data quality, especially by understanding data available outside of clinical trials, such as Compassionate Use/Early Access programs.

10

Food for thought: Your organisation knows how many buildings, computers and even chairs it owns, right? In well-managed companies, all of this is even available at the push of a button. But too often, this is limited to traditional assets, such as raw materials, machines and buildings. Many companies call data an asset. But is your organisation ready to truly treat #DATA as an asset? I'd be happy to read your comments. My bi-weekly "Food for Thought" series invites you to share your ideas. #cdo #datamanagement #foodforthought

21

1 Kommentar

🚨OpenAI Warns About Emotional Attachment to New Voice Feature🚨 During testing of GPT-4o, OpenAI researchers observed users expressing emotional connections with the AI, using phrases like “This is our last day together.”  🤖❤️ Additionally, a recent paper from Google DeepMind also highlights how chatbots’ language abilities can create an impression of genuine intimacy. 🤖❤️ While this might seem far-fetched, anthropomorphizing AI is a real, near-term risk (unlike fantasy AI doomerism). This concern extends to life sciences, where overuse in clinical trials could lead patients to confuse AI interactions with genuine human care, potentially impacting trial outcomes and well-being. Similar risks exist in telemedicine, where patients might become dependent on AI-driven consultations, and in mental health apps, where AI could blur the lines between real human support and machine interaction. I'm not sure of any real solutions apart from humans learning to differentiate between a mechanical interaction 🤖 and a soulful interaction 💖. 😊 #AI #OpenAI #MentalHealth #Telemedicine #ClinicalTrials #AIethics #TechRisks #FutureOfAI #Innovation #genai #lifesciences #pharmaceutical

11

2 Kommentarer

My Friday schedule can be unpredictable - sometimes it's full of excitement, other times it's calming wind-down to the end of a week. This afternoon, however, I had an energizing rumination with Daniel Kalamarides, PhD on the future of #GenAI in healthcare. Good questions always get the wheels in my head spinning. Reflection on our discussion: ✋🏼 I think #prevention is the part of #healthcare where GenAI can be most valuable in. There's never enough investments in prevention vs treatment, but the greatest number of people can benefit. For example, educating, motivating and nudging people at high risk of developing #diabetes, #heartdisease, #cancer, #falls, etc. to change behavior, so we can catch people before they become #patients. 💸 But who will fund GenAI initiatives for prevention? (which stakeholder's interest is aligned with prevention?) Government is an obvious answer, but we all know the dilemma with government budgets. NGOs? Possible. Maybe #vaccine manufacturers? #consumerhealth companies? #Payers / #healthinsurance companies? #Employers? How can #innovators be better incentivized to develop GenAI solutions for prevention, with a sustainable commercialization pathway, so it's not just another one-and-done initiative? ⚔ GenAI is a double-edged sword for patient experience. While we are cautious not to take away the human experience that people are familiar with, being able to speak freely to a non-human may remove the barrier for some people who fear of being judged. What if this can give us insights into the real reasons for non-adherence to medication, for no-shows to medical appointments, or for misunderstandings to medical advice? We still have long ways to truly realizing the value of GenAI in healthcare, but this emerging #digitalhealth technology carries so much promise that gives us permission to dream what can be possible. (image below AI-generated by #Microsoft Designer) #generativeAI #healthcareinnovation #behaviorchange #patientexperience #medicationadherence #perspective #dreambig #shootforthestars #possibilities #whatif

53

8 Kommentarer

What can industry leaders learn from Apple’s approach to AI? Stop using terms like “we are AI first” or “AI led” (What does that even mean in an enterprise context? 😀) Start designing an “AI in everything” plan. AI should be integrated into every single project in the company: SAP upgrades, Sharepoint updates, MES deployment, business led R&D transformations, Veeva deployments, CRM builds… it doesn’t matter! AI in everything is the only way to truly change how business is done. This is especially true for life sciences leaders. A knowledge-rich industry like pharmaceuticals could greatly benefit from AI. Imagine integrating AI into every phase of drug development and operations: optimizing clinical trial designs, predicting patient outcomes, personalizing treatment plans, automating regulatory compliance, intelligently automating core functions, enhancing supply chain management…. Think of every project as an AI project. #AI #ArtificialIntelligence #Innovation #PharmaTech #LifeSciences #DigitalTransformation #AIinBusiness #FutureOfWork #AIinPharma #EnterpriseAI #pharmaceutical #clinicaltrials #IT #supplychain

15

Great article how we can become AI augmented leaders. AI is a powerful tool when we apply it well. The challenge for us as leaders is to leverage it in combination with our compassion, so we can truly drive meaningful change.

27

6 Kommentarer

Fully agree with Paul Hudson on this one! As an industry, we need to become more clear and purposeful on where and how we expect AI to deliver business impact and customer value in the short-, mid-, and long-term. It has to be worth the significant investment of resources and inevitable disruption to our ways of working. In my own team, we are still mostly focusing on applied (or snackable) AI, which is a great starting point for a majority of people. Explorative (or expert) AI is definitely also looked at, and we see the opportunities, but we aren’t prioritizing those use cases until we have our “AI Basics” set up. #backtobasics

83

8 Kommentarer

We need to learn to delegate tasks to AI. Saying no to AI is not an option. The more promising approach is to pave the way for employees to use AI in a compliant and productive manner within the company. This way, AI can take on many time-consuming tasks, significantly increasing the productivity of both people and processes, and ultimately enhancing the quality, innovativeness, and delivery reliability of products. The shift towards #artificialintelligence cannot be stopped. The dynamism and speed at which new AI applications are being brought to market is enormous, and the expectations associated with them are immense. Ultimately, AI is a crucial tool for remaining competitive. Inspiration on how implementation can look within a company is provided by the #podcast "Five Best Practices to Succeed with #Data and GenAI: Lessons from Leaders." Walid Mehanna, Chief Data & AI Officer at Merck Group, also shares his perspective. https://lnkd.in/eec7yWGx

46

1 Kommentar

👉🏼 Performance of a Proprietary Large Language Model in Labeling Obstetric Incident Reports 🤓 Jeanene Johnson 👇🏻 https://lnkd.in/e_8FwVWE 🔍 Focus on data insights: - 📊 The proprietary LLM achieved an overall sensitivity of 85.7%, indicating strong performance in identifying relevant incidents. - ✅ Specificity reached 97.9%, showcasing the model's ability to correctly identify non-incidents. - 📈 The model produced 79 labels in total, compared to the 49 labels created by clinicians, demonstrating a higher labeling capacity. - 🔍 The positive predictive value was 53.2%, suggesting room for improvement in distinguishing true positives from false positives. 💡 Main outcomes and implications: - 🎯 High sensitivity and specificity indicate that the LLM could play a crucial role in improving patient safety protocols through efficient incident reporting. - ⚙️ The results suggest that integrating LLMs into clinical workflows could lead to enhanced data utilization in healthcare settings. - 🔄 Potential for LLMs to automate and streamline the labeling process, enabling healthcare professionals to focus on other critical areas of patient care. 📚 Field significance: - 🌐 Incorporating LLMs into healthcare could revolutionize how incident reports are analyzed, leading to improved patient outcomes. - 💻 This research highlights the need for further exploration of machine learning applications in clinical safety and quality improvement. - 🧠 It underscores the importance of balancing technological advancements with clinician oversight to maintain accuracy and safety standards. 🗄️: [#LargeLanguageModel #PatientSafety #IncidentReporting #HealthcareData #MachineLearning #ClinicalEvaluation]

1

Unlock the full potential of your data with our expertise. We’ll guide your organization in building a data-driven strategy for growth and competitive advantage.

2

Do not miss this great opportunity to learn more about Generative AI in healthcare. I look forward to seeing all of you tomorrow #AI, #ML, #genAI Generative AI for Healthcare

21