The integration of Large Language Models (LLMs) into medical applications has sparked widespread interest across the healthcare industry, from drug discovery and development to clinical decision support, assisting telemedicine, medical devices, and healthcare insurance applications. This perspective paper aims to discuss the inner workings of building LLM-powered medical AI applications and introduces a comprehensive framework for their development. We review existing literature and outline the unique challenges of applying LLMs in specialized medical contexts. Additionally, we introduce a three-step framework to organize medical LLM research activities: 1) Modeling: breaking down complex medical workflows into manageable steps for developing medical-specific models; 2) Optimization: optimizing the model performance with crafted prompts and integrating external knowledge and tools, and 3) System engineering: decomposing complex tasks into subtasks and leveraging human expertise for building medical AI applications. Furthermore, we offer a detailed use case playbook that describes various LLM-powered medical AI applications, such as optimizing clinical trial design, enhancing clinical decision support, and advancing medical imaging analysis. Finally, we discuss various challenges and considerations for building medical AI applications with LLMs, such as handling hallucination issues, data ownership and compliance, privacy, intellectual property considerations, compute cost, sustainability issues, and responsible AI requirements.

The staggering pace with which the capabilities of large language models (LLMs) are increasing, as measured by a range of commonly used natural language understanding (NLU) benchmarks, raises many questions regarding what "understanding" means for a language model and how it compares to human understanding. This is especially true since many LLMs are exclusively trained on text, casting doubt on whether their stellar benchmark performances are reflective of a true understanding of the problems represented by these benchmarks, or whether LLMs simply excel at uttering textual forms that correlate with what someone who understands the problem would say. In this philosophically inspired work, we aim to create some separation between form and meaning, with a series of tests that leverage the idea that world understanding should be consistent across presentational modes - inspired by Fregean senses - of the same meaning. Specifically, we focus on consistency across languages as well as paraphrases. Taking GPT-3.5 as our object of study, we evaluate multisense consistency across five different languages and various tasks. We start the evaluation in a controlled setting, asking the model for simple facts, and then proceed with an evaluation on four popular NLU benchmarks. We find that the model's multisense consistency is lacking and run several follow-up analyses to verify that this lack of consistency is due to a sense-dependent task understanding. We conclude that, in this aspect, the understanding of LLMs is still quite far from being consistent and human-like, and deliberate on how this impacts their utility in the context of learning about human language and understanding.

Relation extraction (RE) is the core NLP task of inferring semantic relationships between entities from text. Standard supervised RE techniques entail training modules to tag tokens comprising entity spans and then predict the relationship between them. Recent work has instead treated the problem as a \emph{sequence-to-sequence} task, linearizing relations between entities as target strings to be generated conditioned on the input. Here we push the limits of this approach, using larger language models (GPT-3 and Flan-T5 large) than considered in prior work and evaluating their performance on standard RE tasks under varying levels of supervision. We address issues inherent to evaluating generative approaches to RE by doing human evaluations, in lieu of relying on exact matching. Under this refined evaluation, we find that: (1) Few-shot prompting with GPT-3 achieves near SOTA performance, i.e., roughly equivalent to existing fully supervised models; (2) Flan-T5 is not as capable in the few-shot setting, but supervising and fine-tuning it with Chain-of-Thought (CoT) style explanations (generated via GPT-3) yields SOTA results. We release this model as a new baseline for RE tasks.

This acquisition will further expand its suite of capabilities that deliver an enhanced customer experience and security. Artificial intelligence plays a critical role in enabling consumer convenience, while delivering enhanced security. This advanced technology delivers greater insights from every transaction to assist in making even more accurate fraud decisions. "To fully realize the promise of our increasingly digital lives, we need to design our payment systems with the future in mind and that's what we're doing," said Ajay Bhalla, president of enterprise risk and security for Mastercard. "Our unprecedented use of artificial intelligence on our network is already proving successful. With the acquisition of Brighterion, we will further extend our capabilities to support the consumer experience."

The solution uses artificial intelligence technology to help financial institutions increase the accuracy of real-time approvals of genuine transactions and reduce false declines. This is the first use of AI being implemented on a global scale directly on the Mastercard network. Current decision-scoring products are focused primarily on risk assessment, working within predefined rules. Decision Intelligence is a radical new approach that goes much further. It takes a broader view in assessing, scoring and learning from each transaction.

MasterCard (MA) is taking payments, the last -- and sometimes, the most painful -- part of shopping into the future, replacing card swipes with robots, artificial intelligence and the ubiquitous selfie. The steps, including this week's introduction of a chatbot for banks called "MasterCard KAI" that uses artificial intelligence to respond to customer queries via texts or through apps like Facebook Messenger, are vital parts of CEO Ajay Banga's strategy of leverage technological development to expand the $113 billion company beyond traditional card-based transactions. To create the chatbot, MasterCard partnered with startup Kasisto, developing a "conversational artificial intelligence platform" that banks and merchants can use to let customers make transactions, monitor their spending habits, check account balances and ask questions, the company said at the Money 20/20 conference in Las Vegas. It will be released in the U.S. early next year. "This bot enables entirely new experiences, bringing Mastercard benefits and offers to consumers with human-like conversations that are personal and contextual," Zor Gorelov, Kasisto CEO and co-founder, said in a statement.