It is nearly impossible to read any business news without running across the new ways that the OpenAI® ChatGPT® 2023, that is being used by both professional and non-professionals, in many fields. This is making people from schools to boardrooms think of its potential applications. We see this as an inflection point in both the development of the technology and its use by nearly all Americans. We even used it in our last blog and asked it, “What are the financial and operational challenges faced by US physicians in 2023?” The response we posted, without any editing in the blog, and frankly it was right on point.
Those that are fanatics of science fiction recall what we would today call uses of AI in book and movies. First to come to mind is the television series “Star Trek” and its “computer” that was always available to provide information and analysis to nearly every situation. Some of my friends have even renamed their Amazon Alexa as “computer” in homage to Star Trek. Some of us may also remember the 1968 movie “2001 A Space Odyssey” based on Sir Arthur C. Clarke’s novel and the infamous HAL 9000, a Heuristically programmed algorithmic computer is a sentient artificial general intelligence computer. We can also think of robot with AI like in The Robot from the series Lost in Space (1965-1968), or if we get really dark “Terminator” with Arnold Schwarzenegger or “iRobot” with Will Smith.
We have finally arrived at the future.
Background on new AI platforms
The AI driven platforms are being developed as fast as apps were being developed for the iPhone 10 years ago – seems like one a day almost. There are some developed for ‘general use’ (like ChatGPT) that claims to generate blogs and papers, and there are AI apps developed by OpenAI, a company that Elon Musk helped organize originally as a non-profit. This is the way ChatGPT describes itself:
“an AI-powered chatbot developed by OpenAI, based on the GPT (Generative Pretrained Transformer) language model. It uses deep learning techniques to generate human-like responses to text inputs in a conversational manner.”
ChatGPT is driven by a Large Language Model or LLM, that allows it to understand human language (to an extent) and based on Natural Language Processing (NLP) to generate a reply to questions or queries using its access to data. Today, ChatGPT has over 150 billion parameters, and clearly the largest NLP capacity ever invented.
Actually, using it is incredibly easy and the responses surely appear human like. The opportunity to edit and adjust questions or queries allows a response to be ‘fine-tuned’ for greater accuracy. One of the most convenient features is the ability to upload information for it to use or re-write from its own database.
Open to the public in December of 2022, it quickly amazed 100 million global users in only two months.
Though ChatGPT is not yet done, (there is a new enhanced version soon to be released called ChatGPT-4) original entry into the market has proven the value of AI to millions.
The pinnacle of AI is generally referred to as Artificial General Intelligence, or AGL, and that refers to the ability of intelligent agent, that can learn, compare entities like two people, and make differentiating decisions between them and also perform most tasks a human can do. In that race, you have companies in the running, one that quickly comes to mind is Google’s DeepMind®.
So how WILL all this impact healthcare?
Once the dust settles, we believe that amazing opportunities will emerge. AI, when properly configured, can operate at blazing speed. This is both a benefit and nemesis because without proper design, it would be putting an 18-year-old in a new Ferrari with $100M free money. That is a recipe for certain disaster; however, this same 18-year-old, when properly educated and trained, given boundaries and clear objectives could do a lot of good!
We know that the universe of data available in healthcare alone far exceeds the ability of any one person to accumulate, aggregate, digest or analyze, and disseminate properly. This is where AI can help by making “learned” decisions about what is germane to a clinical question and what is not. For instance, a question regarding a patient’s history regarding a cardiovascular issue does not need data aggregated regarding an orthopedic issue.
The impact of AI on healthcare will be in multiple areas, but first and foremost in synthesizing the volumes of data, both clinical and patient specific to the extraordinary libraries of knowledge base on medical research. To be able to slice through the dozens of data silos to extract and synthesize data on a given subject would be an incredible advantage in achieving our fundamental corporate objectives of:
- empowering the patient, employers & public health agencies,
- enabling the provider,
- engaging the payor, and
- advancing the science of healthcare.
HealthScoreAI our latest launch is utilizing the first seven (7), or a portfolio of 34 issued patents to chart these unknown waters. There are some who are focusing on niche situations like diabetes research, and we applaud these early-stage efforts.
Later, we see AI becoming the physician’s clinical assistant by analyzing and evaluating the patient using tools we have today including telemedicine, remote capture of patient vitals, measuring voice and physical changes between visits, and performing comparison of the patient to other patients whether in the physicians’ care or not. These are activities that will produce a much higher quality of care, as well as reduce the overall cost of care while reducing the physician-patient cycle time. The time the physician spends with the patient, when they are loaded with more than current vital signs and a complaint, will astronomically increase the accuracy of diagnosis.
The effect of having a trained “HealthConsultant” to perform all these tasks enables both the provider and empowers the patient far beyond anything we have or are doing today. This is the future of healthcare – a future that can be as effective remotely as in the physician’s office.
We’re excited about the future of healthcare today and can see a major pivot in just a few years ahead. We don’t forecast a robot taking over a patient’s care despite the reports of ChatGPT passing a medical bar exam. The practice of medicine is far more than knowledge, it is about wisdom and humanity, and that comes with experience as a doctor and as a human that cannot be easily distilled into a digital file.
More to follow.
-Noel J. Guillama, Chairman