Understandably, there is considerable focus on how GenAI may impact business. Who is making the money and how they are making it. But much more profound effects are likely to be felt in our daily lives, including where GenAI enables us to uncover new frontiers in science and medicine.
Recommended Reading
AI tools are becoming increasingly common in science, with more than 1,600 researchers anticipating their importance in the next decade. However, concerns about AI's impact on research include reliance on pattern recognition without understanding, bias in data, fraud, and irreproducible research. Large language models (LLMs) like ChatGPT are both impressive and concerning, with worries about misinformation, plagiarism, and mistakes in research papers. Barriers to AI adoption include lack of resources, skilled scientists, and experience. Commercial firms dominate AI resources, and collaboration with them is seen as important. Reviewing papers that use AI is a challenge, and concerns about AI's impact on society include spreading misinformation and AI-assisted surveillance.
AI is industrializing biopharma and healthcare in everything from drug design and diagnostics to healthcare delivery and back-office functions. Bio can now scale, massively.
The article discusses Enterprise Large Language Model (LLM) projects, emphasizing their varying impacts across categories like data transformations, natural language interfaces, workflow automations, copilots, and autonomous agents. It highlights the significance of workflow automations for business value, provided the right candidates are chosen based on specific criteria. The potential of chatbots and copilots is acknowledged, but autonomous agents are seen as the ultimate productivity boost, requiring substantial resources. The article concludes by hinting at a forthcoming installment addressing poor use-case choices and risks in LLM projects.
Large language models have exhibited exceptional performance on various
Natural Language Processing (NLP) tasks, leveraging techniques such as the
pre-training, and instruction fine-tuning. Despite these advances, their
effectiveness in medical applications is limited, due to challenges such as
factual inaccuracies, reasoning abilities, and lack grounding in real-world
experience. In this study, we present ClinicalGPT, a language model explicitly
designed and optimized for clinical scenarios. By incorporating extensive and
diverse real-world data, such as medical records, domain-specific knowledge,
and multi-round dialogue consultations in the training process, ClinicalGPT is
better prepared to handle multiple clinical task. Furthermore, we introduce a
comprehensive evaluation framework that includes medical knowledge
question-answering, medical exams, patient consultations, and diagnostic
analysis of medical records. Our results demonstrate that ClinicalGPT
significantly outperforms other models in these tasks, highlighting the
effectiveness of our approach in adapting large language models to the critical
domain of healthcare.
Disparate sources of unstructured data, which abound in healthcare, are now assets to power generative AI. Here’s how private payers and healthcare systems can use the technology to deliver care.
Authors: Justin Norden, Jon Wang, and Ambar Bhattacharyya
We have written dozens of articles about generative AI already, covering all possible angles, but just realized that we have never covered the basics, and explained what is generative AI in the first place. Generative AI refers to a category of AI algorithms that look for patterns and structures in
Northwell Health, New York's largest health system, is strengthening its commitment to generative AI technology through a partnership with Aegis Ventures. The collaboration, backed by $100 million in joint venture funding, focuses on using generative AI models to streamline administrative tasks for healthcare providers. The initial goal is to reduce the time doctors spend on paperwork, enabling them to spend more time with patients. Generative AI models can create human-like text and content based on data patterns, making them ideal for addressing administrative burdens. The partnership aims to leverage technology to enhance healthcare efficiency and reduce costs.
The Hippocratic Oath, guiding medical ethics for centuries, should be updated to reflect digital health's impact on medicine. Proposed revisions include acknowledging various contributors to medical progress, recognizing preventive care, embracing digital technology's role, validating patient contributions to treatment, addressing data privacy, and emphasizing patient-centered care amid technology advances. These updates would ensure the Oath's relevance in 21st-century healthcare.
Google debuts new Vertex AI Search capabilities for health providers. The new feature allows organizations to build their own generative AI-enabled search engines for customers. It can be combined with Med-PaLM 2 to find answers to medical questions and retrieve accurate clinical information. The offering aims to address workforce shortage, relieve administrative burden, and provide clinical decision support.
Foundation models (e.g., ChatGPT, AlphaFold) have inspired interest in medical records. Reviewing 84 models, we reveal gaps in understanding, propose evaluation improvements, and explore healthcare benefits.
Forward Health has launched CarePods, a self-contained, AI-powered medical station that allows users to perform various medical tests without the need for a doctor or nurse. The CarePods use custom AI for diagnosis, and doctors review the findings and issue prescriptions remotely. The monthly cost is $99, and Forward Health plans to scale the deployment of CarePods to 3200 units within a year. The company also offers traditional tech-forward doctor's offices with full biometric assessments. Forward Health has raised $657.50 million in funding and has an impressive board of directors and advisors from Silicon Valley.