WHY IS ARTIFICIAL INTELLIGENCE IMPORTANT?
Generative artificial intelligence (GenAI) is the master platform for transforming health care. It will remake and modernize health care provision during your career. It is driving both cultural and technological change in health care. This transformation brings many exciting opportunities to improve patient care, especially access and affordability, while increasing revenue. It also offers you the chance to be a changemaker and assume leadership in this unstoppable transformation of hearing health care (HHC). The capabilities and progress of HHC will depend on how well we transform HHC into a Generative (GenAI)-based system while retaining our core values.
THE AI PROBLEM:
AI technology is incredibly powerful and evolving rapidly, fundamentally altering the landscape of HHC. This Intelligence Revolution is disrupting traditional HHC delivery and business models, and the path forward is uncertain. As providers of HHC services, AuDs are at the forefront of this transformation. Your role in this transformation is significant, and your contributions are invaluable.
THE SOLUTION:
The resolution to taming this disruption is understanding GenAI and harnessing its potential to create new AI-enabled healthcare delivery and business models that solve HHC’s most pressing challenges. This guide aims to create that knowledge, encourage the use of AI to solve HHC’s struggles and rethink the assumptions of healthcare delivery.
AI OVERVIEW
Professor John McCarthy created the term artificial intelligence (AI) in 1956 when he gathered a small group to spend a few weeks pondering on how to make machines do things like use language (Simonite, 2023). They failed, but they planted a fertile seed. AI, the ability of software to perform cognitive functions traditionally associated with human minds, became a new field of study. AI has been in our lives for years. When we talk to Siri or Alexa, search the internet or use a chatbot, we’re taking advantage of AI. However, basic AI could not generate original content. That innovation came recently.
Deep-learning-based models use circuits and algorithms based on brain neural networksnested in layers, with connections between and among layers weighted differently as they train and learn. The first layer receives the input, and the last layer yields the output. Just as scientists who study the brain don’t understand precisely how the brain works, the experts who create neural networks don’t always understand what happens in the neural networks they make. Deep-learning models excel at learning from text, images, audio and code. From this information, they can produce new text, images, audio, code, simulation and videos. They can understand sequential data, like how a word is used in a sentence, and drastically change the way we approach content creation.
Researchers at DishBrain are taking neural network modeling to the next level by fusing computer chips with living human and mouse brain tissue. Their goal is to enhance neural network-based AI models with biological intelligence. According to Blain (2023), this approach shows promise.
Machine learning (ML) is a form of AI that can learn from data patterns without human direction. We train ML on an extensive database that detects patterns and learns how to make predictions and recommendations. It also adapts, becoming more efficient with new data and experiences.
Generative AI (GenAI) is a form of machine learning based on deep learning. It can generate new content responding to a prompt by identifying patterns in massive quantities of training data and creating original material with similar characteristics. Outputs from GenAI models can be indistinguishable from human-generated content. GenAI can be used out of the box or fine-tuned to perform specific tasks.
Large language models (LLM) are a type of GenAI, such as ChatGPT, trained exclusively on text. Because language allows us to build models of the world without any other stimuli, like vision or hearing, LLMs can write fluently about the relationships between different sounds even though it has never heard either.
The Intelligence Revolution refers to the massive transformation of society caused by the exponential growth of computer power and the unrelenting desire to create machines that do everything humans do. It is a profound revolution in the way we think, work and view ourselves as humans. It is transforming our society, including the HHC professions.
In later sections of this guide, we will consider how AI facilitates HHC provision, including machine-learning models and their predictions and the new systems for care delivery they enable.
WHAT ARE THE ADVANTAGES OF GenAI?
- GenAI drives down the time taken to perform a task. It enables multitasking and eases the workload for existing resources. These advantages improve productivity and increase cost savings.
- GenAI enables the execution of hitherto complex tasks without significant cost. It sidesteps the need to hire competent but expensive new experts.
- GenAI operates 24/7 without interruption or breaks, surpassing the dedicated performance of even our most loyal and committed clinic employees.
- GenAI facilitates decision-making by employing a wide base of information, making the process faster and wiser.
- GenAI allows for the rapid query, analysis and summary of massive amounts of data. As detailed in later sections of this guide, this enables precision medicine in HHC, an innovativeapproach totailoring disease prevention and treatment that considers differences in people’s genes, environments and lifestyles.
- GenAI is being deployed across industries. It is the fastest-spreading innovation ever. Patients will expect HHC to use GenAI, and its use will define the best care.
WHY IS GenAI CHANGING SO RAPIDLY?
It is a powerful human bias to expect tomorrow to be like today. So, we wildly underestimate how quickly AI systems will transform health care. Let’s look at what is countering our bias and driving the rapid changes in AI.
- GenAI is popular: ChatGPT drew one million users in the first five days of its existence. In 40 days, it had 100 million users. That is the fastest adoption of any innovation in history.
- GenAI is easy to use: It operates with natural language processing, which means it understands instructions in natural language, the language you use daily. There is no need to know computer programming to communicate with or instruct GenAI.
- GenAI knowledge diffuses quickly: Researchers from competing AI labs hang out socially and discuss their work. AI researchers also publish more papers and give more presentations than most scientists.
- GenAI feeds itself: GenAI is self-improving. Now that we can partner with AI, we can improve and amplify what we do to push science and technology forward. This results in a massive increase in scientific and technological advancement, creating a more powerful AI. This leads to greater advances in technology and science, which, in turn, further improves AI. In this feedback loop, AI progresses rapidly.
- GenAI is competitive: Big tech and startups know its promise. They are spending billions of dollars in an epic race for AI platform dominance. In 2023, enterprises spent $16 billion on GenAI solutions. According to the International Data Corporation, they are expected to spend $143 billion by 2027.
- GenAI is universally necessary: Business leaders realize that GenAI is crucial to staying competitive across industries. Companies know they need to add AI to their operations or get left behind. And HHC providers are not exempt from this fact.
- GenAI is not stopping or pausing: The AI race is in full swing, forever changing how we provide HHC. Competitors spend billions of dollars to win the race for AI dominance. The transformation train has left the station, and we cannot stop it.
FUTURE AI:
Here is a look at the advances AI will make going forward.
Artificial general intelligence (AGI), the stage at which AI can do any job that a human can do, only better, is the long-term goal of AI. It is a central theme of the Intelligence Revolution. This guide will not specifically cover AGI, but it is a fascinating, controversial and fast-moving field of study that offers a world without work (Susskind, D., 2020). It raises the supposition that AI may not be artificial. Instead, we should refer to it as inorganic or machine intelligence. We encourage readers to keep abreast of AGI, monitor its progress and imagine how it will affect HHC. But first, this guide provides an essential background of GenAI and its application to HHC. In later sections, we will attempt to answer the fundamental question: What do we do?
Artificial superintelligence (ASI) is a hypothetical future AI that is significantly more intelligent than the best human. ASI does not exist.
Improving AI
There are three main dimensions to quickly improve AI: size, data and applications (Bertics, A., 2023).
- Size: Traditionally, we considered larger models to be better. However, increasing size results in enormous costs. The new focus is to maintain performance by making models smaller and faster. This transformation is achieved by training a smaller model using more data. We can shrink size by reducing the numerical precision of the parameters within a model. Small models are less expensive to run and more accessible.
- Data: We can also shift the focus from data quantity to data quality. Furthermore, we can create more effective models by increasing and combining data types to expand capabilities.
- Application: AI is evolving quickly. One way to improve AI is to learn how to use it more effectively. There are three main ways to use AI:
- Prompt engineering feeds the model with specific phrases or questions based on the desired goal.
- Fine-tuning a model to improve it, including adding an extra round of training using papers from HHC journals to make it better at answering HHC questions.
- Embed LLM in a more extensive, robust architecture, including combining an LLM with extra software and a database of knowledge to make it less likely to produce error-ridden content
Computer Advances and AI
The power of AI comes from its training materials and computer power. As computer power grows, so does AI.
Personal computers (PC): In 2024, PC power is growing due to the addition of GenAI to smartphones and personal computers, allowing them to run GenAI algorithms directly on their hardware without the internet or expensive cloud computing services. Lisa Su, who leads Advanced Micro Devices, says AI-enabled PCs will fundamentally redefine the computing experience over the coming years.
Supercomputers: In one second, Aurora, the newest exascale supercomputer, can perform two quintillion operations (the number two followed by 18 zeros). It will be functional in 2024 and have 70% more memory than the previous top supercomputer. Aurora’s creators will equip it with the latest advances in AI and use it to address medical issues, among other goals. Lawrence Livermore National Laboratory and Tesla are each building their own powerful supercomputers.
Quantum computing has the advantage of being quantum-based, like nature, rather than digital-based, like most supercomputers. This means that it can simulate the reality that digital computers struggle with. The best supercomputer, before Aurora, would take an astonishing 47.2 years to match a computation by Google’s newest quantum computer (Kaku, M., 2023).
The rapid growth of more powerful computers and the accompanying expansion of expertise will accelerate the use and capabilities of AI. Improved AI can help design even faster, more powerful computers. The Intelligence Revolution is on a fast track!
Integrating novel data and providing new services is crucial to healthcare AI development. As GenAI combines with HHC, we must shape it to solve critical problems in the HHC environment.
Resources for Further Study
The Future of the Professions: How Technology Will Transform the Work of Human Experts, Updated Edition
By Richard Susskind, and Daniel Susskind, (2022). Oxford University Press
This book is a must-read if you want to know how AI will impact audiology. It is the first book to ask if professions still matter in the 21st century. It predicts how professions will decline and introduces the people and systems that will take over. We won’t require or desire doctors, teachers and other professionals to work the same way as in the 20th century. In this transformation era, it’s on us to redefine our profession. What to do, who does it and what not to do. A basic understanding of AI and its practical implementations is vital for effectively revolutionizing audiology and HHC. This book explains how technology will transform professions. It will inspire you to consider in depth how audiology will be transformed.
The Intelligence Revolution in Hearing Health Care Delivery
By Donald Nielsen, (2024). A Fuel Medical Group Publication.
Available at https://fuelmedical.com/wp-content/uploads/2024/10/fm_march2024_intelligence_revolution_paper_v1.pdf
Much of the content in this piece is derived from this publication. Dr. Nielsen provides a comprehensive analysis of AI, detailing its impact on hearing healthcare delivery. He describes the transformative impact of AI on our healthcare system and briefly introduces precision medicine and genomics.
The intelligence revolution presents countless opportunities for HHC providers to grow and expand. The paper argues in favor of incorporating AI and precision medicine into HHC, stressing the significance of holistic care, continued learning, ethical AI implementation and cooperation among diverse healthcare sectors.
If you are new to AI, precision medicine or genomics, this paper is an excellent introduction. The summary covers the evolution of HHC from the past to the present and looks ahead to the future and the healthcare revolution that will benefit us going forward.
Co-Intelligence: Living and Working with AI
By Ethan Mollick, (2024). Portfolio Penguin
Wharton professor Ethan Mollick is a prominent and provocative scholar on AI, focusing on the practical aspects of how these new tools for thought can transform our world. In Co-Intelligence, Mollick urges us to engage with AI as co-workers, co-teachers and coaches. He assesses its profound impact on business and education, using dozens of real-time examples of AI in action. By following his lead, you can develop your own examples of how it can be used in HHC. Co-intelligence shows what it means to think and work together with smart machines and why we must master that skill. Mollick challenges us to use AI’s enormous power without losing our identity, learn from it without being misled and harness its gifts to create a better human future. Let’s accept his challenge.
The AI Revolution in Medicine: GPT-4 and Beyond
By Lee, P., Goldberg, C., Kohane, I., (2023). Pearson Education, Inc.[VH2]
This book is the ultimate opportunity to learn about GPT-4’s use in healthcare from three insiders who have had exclusive access to it for a while. They show how it can significantly improve diagnoses, summarize patient visits, streamline processes, speed up research and much more. Brace yourself for authentic GPT-4 dialogues, no rehearsals or filters, the good and the bad, with valuable context, honest commentary, real risk insights and up-to-the-minute takeaways.
References
Bertics, A., (2023) The Economist Nov. 13, 2023. This article appeared in the Science and Technology section of the print edition of The World Ahead 2024 under the headline “What’s next for AI research?” p 91-92
Blain, L., (2023). New Atlas Jul. 21, 2023, Computer chip with built-in human brain tissue gets military funding, available at :https://newatlas.com/computers/human-brain-chip-ai/
Kaku, M., (2023) Quantum Supremacy, Double Day, New York
Simonite, T., (2023). The Wired Guide to Artificial Intelligence, Wired Feb. 8, 2023. Available at: https://www. wired.com/story/guide-artificial-intelligence/
Susskind, D., (2020) A World Without Work: Technology, Automation and How We Should Respond, Metropolitan Books, Henry Holt and Company, New York.