MedpaLM: The Doctor of Tomorrow
How Google’s medical-oriented large language model is recasting the healthcare industry
By Shehraan Hafiz
Imagine a world without doctors. While it might sound like a dystopian nightmare right now, it could also be a vision of the future. Technology has made great advancements in recent years and this has allowed it to have the potential of taking over the role of diagnosing diseases. AI is already being used to assist doctors in various tasks, such as analyzing medical images, detecting anomalies, recommending treatments, and monitoring patients. AI could potentially replace doctors in some scenarios, such as providing primary care, performing surgery, and conducting research. However, what about AI could enable this feat?
Introduction to AI
Artificial intelligence (AI) refers to the utilization of computers and machines to mimic human intelligence and execute intricate tasks. In the realm of healthcare, AI holds immense promise for enhancing various aspects, including diagnosis, treatment, research, and administrative processes. In this overview, we’ll delve deep into MedpaLM, Google’s in-house implementation of AI as a large language model, as well as examine the advantages and hurdles associated with this transformative technology.
But what is a Large Language Model (LLM)?
A LLM is a deep learning algorithm that can perform a variety of natural language processing tasks. So might be wondering, what exactly is deep learning? Well, deep learning is perhaps the most complex subfield within machine learning so I will be using a simple analogy to help explain it
Imagine this: You want to create a simple program that can categorize between different fruits, like “orange,” “apple,” “mango” etc. Deep learning algorithms would figure out the features of each fruit to distinguish them from each other. For example, it could use a feature like the object being an “orange sphere,” to identify that the fruit is an orange “orange“ or that “Spongebob’s underwater home” is a “Pineapple” (that last one was a joke).
But now, your next question might be “What is natural language processing (NLP)?” In simple terms, NLP is a machine learning technology that gives computers the ability to interpret, manipulate, and comprehend human language. It is primarily used in chatbots, voice assistants and translation applications. If you ever talked to Alexa or Siri, then know that NLP is what is enabling the computer to speak to you.
What is MedpaLM?
MedpaLM uses the aforementioned Large Language Model concept to provide high quality answers to medical questions. It is fine-tuned for the healthcare industry and has been tested on medical exams, medical research, and customer queries. Unlike many other LLMs, MedpaLM is completely open-sourced with the MIT license, meaning anyone is able to view its source code.
MedpaLM‘s achievements
- Passes the U.S. Medical Licensing Examination (USMLE with over a 60% mark.
- Reaches a 86.5% accuracy on the MedQA medical exam benchmark in research
- The newest model, MedpaLM 2, is also able to work with everything from images, sensors, wearables and more
PaLMedic: Your medical AI chatbot
Now that I’ve introduced MedpaLM to you, don’t you just wish that you could see it in action for yourself?
Well congratulations because now you can, all thanks to my latest program, PaLMedic.
PaLMedic allows anyone to quickly ask questions to MedpaLM in a few simple steps. Check out my Youtube video below to see how it all works:
Please note that the program is currently only available for use in certain regions. Check out the program here, to discover if you live in one of these regions
Hey, I’m Shehraan, a 17-year-old driven to impact the world using emerging technologies. If you enjoyed my article or learned something new, feel free to follow me on Medium to keep up with my progress in Artificial Intelligence exploration and an insight into everything I’ve been up to. You can also connect with me on LinkedIn! Thank you so much for reading.
