Dan Burnett

I'm thrilled to announce significant upgrades to DocTrain, the Virtual Patient GPT, designed specifically for advanced medical training. This cutting-edge tool now includes: 1. Real-time Health Data Analysis: Harness live data to enhance diagnostic accuracy. 2. Advanced Diagnostic Tools: Utilize AI-driven tools to assist with diagnosing complex conditions. 3. Interactive Case Studies: Engage with detailed scenarios to hone your clinical decision-making. 4. Voice Interaction Capabilities: Train using voice commands for a more intuitive learning experience. 5. Multilingual Support: Access training in multiple languages, making it more accessible globally. 6. Integration with Medical Databases: Connect seamlessly with a vast array of medical knowledge resources. 7. Personalized Learning Pathways: Customize your learning experience to match your training needs. 8. Virtual Reality Integrations: Immerse yourself in hyper-realistic training environments. 9. Ethical Decision-Making Scenarios: Explore complex ethical dilemmas in a controlled, educational setting. 10. Peer Collaboration Features: Collaborate and discuss with peers to enhance learning outcomes. DocTrain is committed to transforming medical education by providing a dynamic and interactive training environment that is accurate, engaging, and professionally enriching. Stay tuned for our upcoming webinars and live demos to see DocTrain in action! #MedicalEducation #HealthTech #AIInHealthcare #MedicalTraining #VirtualPatient

4

2 Comments

John Snow Labs #Healthcare #NLP & #LLM 5.3 is now out! This release completely overhauls entity resolution - mapping medical entities in text to standard medical ontologies, in context - greatly improving accuracy. Full release notes: https://lnkd.in/gsiGF26i Install software: https://lnkd.in/gsTqkWKT #ai #generativeai #healthcareai #healthai #datascience #ontologies #medicalcoding

39

1 Comment

University of California, San Francisco Chancellor, Sam Hawgood, in an interview with Matthew Lungren MD MPH at #ConvergeSummit has outlined the use of #AI scribes to create clinical notes as well as other #AI-driven technologies. Key to their success: a group of practicing physicians with knowledge of AI ensuring that the technology fits into the physicians’ existing workflow, which improves physician adoption of the technology. Another key: measure success through reduction in physician burnout and improvement in patient satisfaction, not increases in physician productivity. Final key: have a real-time, frequent review of implemented technologies to ensure that they still support and are relevant to changing clinical care standards - and shut the technologies down if they are not supporting their purpose.

10

𝐇𝐮𝐦𝐢𝐥𝐢𝐭𝐲: 𝐓𝐡𝐞 𝐒𝐞𝐜𝐫𝐞𝐭 𝐈𝐧𝐠𝐫𝐞𝐝𝐢𝐞𝐧𝐭 𝐭𝐨 𝐒𝐮𝐜𝐜𝐞𝐬𝐬 🚀 “No task is beneath me.” “I used to clean toilets.” Nvidia CEO Jensen Huang shared this with a Stanford class, stunning everyone. Nvidia, now worth $3.32 trillion, started with humility. Building a successful business goes beyond smart people, hard work, and luck. It’s about staying humble, doing the small tasks, and valuing every effort. Humility isn’t just a trait; it’s a game-changer. It shapes leadership and drives true success. #Leadership #Humility #Success #Entrepreneurship #BusinessTips

3

Langgraphs and multi agentic flow with contextual memory can enable the best customer service. We see a future where the Lang graphs deep technology over open source platforms can enable data structuring and even a better informed ai assistants throughout the spectrum. #langchain #crewai #llm #langgraph #andrewng DocsTribe AI

3

Using #GenAI tools to improve #healthcare is a positive move. This published example from a collaboration between Dell Technologies and Northwestern Medicine shows great promise using Multimodal LLMs! Lots of detail to be consumed in this blog and the linked study report. https://lnkd.in/gaGkhWsA

36

This is fantastic tool that enables you to effortlessly compare LLMs and providers based on factors such as cost, speed, and various quality parameters like context window size.

24

In this 9-minute episode of the Beyond Electric Sheep Podcast on AI, Brick Thompson and I discuss how Brick asked Anthropic’s Claude 3.5 Sonnet to write the code for a Pong game. It did so and, after a few requests for modifications, produced a usable game. This capability has huge implications for software developers, and for the evolution of AI. I asked Brick if he would advise a young person to pursue a programming career given AI’s new code-writing capabilities. Brick concluded that, for the mid-term future, there will still be opportunities for work in requirements definition, quality assurance, and security/safety. In the longer term, the best advice is to become an AI engineer. https://lnkd.in/gNkS8Bvg #AI #Software Development #Claude #Anthropic #Career Planning

3

Thank you Mr. Moore, I appreciate the succinct (5 short paragraphs) outlining the 3 critical narratives around AI. Very accurate from my perspective and helpful to keep in mind as the hype and noise only heightens around AI and all it’s potential bounty and failings. All the best.

1

Memia 2024.29: unpicking the aftermath of last week's CrowdStrike incident , anticipating AI arms race dynamics in amongst the torrent of other AI and emerging tech developments.

4

#AI for #Terraform Few months ago, we released the first AI that automatically generate #CloudArchitecture/templates from a simple prompt, making design effortless. We called that AI, bob. But AI can do much more for the #cloudcomputing industry; 🔄 Enhance the import process with improved code and suggestions. 🖥️ Edit your architecture directly through a command prompt interface. 💡 Get tailored configurations and settings for ID cards within the system. 📊 Receive live updates on any architecture issues for quick troubleshooting. 📚 Search through Terraform documentation with ease, saving time. 💬 Understand or update your architecture by chatting with the code itself. ✏️ Turn hand-drawn sketches into structured architectural designs. 🧩 Include customer-specific modules to enhance personalization. 🤖 Get automatic resource suggestions based on your actions. ⚙️ Pre-populate resources and node configurations for a quick setup. 🔄 Convert architectures across different cloud service providers. 📝 Start your designs with a basic whiteboard-like tool. 💻 Get AI assistance in generating or improving #infrastructureascode code. 📈 Enhance documentation quality and visualize your architecture. 🔒 Integrate error checking and security assessments into your CI/CD pipeline. 🖼️ Automatically create visual representations of your cloud architectures. What are you excited about the most?

22

1 Comment

Testing my uncensored LLM. It uses any of several widely available models and can run on Nethena20b, Psionic Catacean, Or Wizard-Vicuna just to name a few. She runs locally on my desktop using Faraday, or WebUI, or even just Kobold. There is no question it won't answer or task it won't help with. It has literally no morals. It's not for sale - I mostly use it to write dark jokes. Still think you can regulate your way around malicious chatbots? Criminals, corporations, and foreign adversaries will never follow those rules. You're more likely to impair researchers and fiction writers than deter malicious actors. Focus on punishing large entities who abuse LLMs rather than regulating how LLMs behave in general. Regulate the people and the companies, not the machines. #AI #LLMs #Tech #Future #Censorship #Safety #Regulation #FreeSpeech #FreeThought

4

1 Comment

New "Iontronic Memristor" Could Revolutionize Brain-Like Computing Using Only Salt and Water. Scientists from the Netherlands and South Korea say they have made a significant leap towards artificially replicating the human brain by developing a device called the “iontronic memristor.” Researchers say this novel device, barely larger than a human hair, functions as an artificial synapse, mimicking the brain’s unique ability to process and share information - https://lnkd.in/gn-5XzCU #AI #memristor

8

1 Comment

There's an interesting debate happening right now in AI about whether Large-scale Language Models and transformer-based architectures will get us to AGI, or whether a different approach is required. On the one hand, proponents of LLMs argue that by increasing the scale of these models, along with some engineering tweaks, we can reach AGI or even superintelligence. However, this approach has many downsides. For one, scaling larger models means training on more and more data. The state-of-the-art models have already been trained on the internet, and companies like OpenAI are now acquiring private datasets to scale their models further. In addition, these models are power-hungry, too, requiring large data centers with hundreds of thousands of GPUs. It's estimated that 47GW of new power generation capacity will be needed by 2030 solely for data centers. The other downside of LLMs is that they can't learn new information like humans do. You need to feed the LLM all the context it needs to carry out a task whenever you call it. The standard workaround for this is storing a working memory in an external memory store like a database and only providing the most relevant memories to the LLM whenever we call it. However, this approach has limitations because you are limited to how many bits of memory you can retrieve and provide to the model based on its context window. So, what are the alternatives to LLMs? Researchers like Yan LeCun, the godfather of deep learning, challenge the idea that LLMs are the pathway to AGI, arguing that these models are essentially stochastic parrots that mimic intelligence without reasoning ability. Human brains develop very differently from LLMs. As babies, we are not born with the whole internet in our brains! Instead, we can learn quickly, understand new concepts on the fly, and put them into practice, like walking, reading, or solving algebraic equations. Researchers who are bearish on LLMs advocate for alternative architectures and approaches, such as objective-driven systems that create a "world model" they can use to predict the outcomes of various actions and plan specific steps to achieve a predefined objective. Other approaches involve using a hybrid of LLMs trained on text with other data types like speech, video and images. This is essentially what OpenAI has done with GPT 4o. Over one trillion dollars will be spent on training AI models in the coming years, so this is a debate to watch closely. To learn more, check out the podcast episode linked in the comments.

16

8 Comments

There's no need to manually put together site reports. With this Python script, you can pull data sent to ODK Central into a customized Word document in seconds. If you aren't yet comfortable running a Python script, you can do the same thing with our drag-and-drop Zapier integration at https://lnkd.in/gxZvFm3q.

14

Excited to highlight an innovative company leveraging AI to transform cancer care: Color Health. In collaboration with OpenAI, they've developed a GPT-4o powered copilot that's revolutionizing how healthcare providers approach cancer screening and treatment. Key impacts: - 4x more efficient identification of missing diagnostics - Reduction in patient record analysis time from weeks to just 5 minutes - Personalized screening plans based on individual risk factors As a tech leader, it's inspiring to see AI being used to improve patient outcomes and potentially save lives. This is a powerful example of how technology can address critical healthcare challenges. What are your thoughts on AI's role in healthcare? How do you see it shaping the future of patient care? Learn more: https://lnkd.in/ekbfk4bc

1