Large Language Models and Responsible AI Define this Year’s NLP Summit

Updated on September 29, 2023

Fall is upon us, bringing with it a bevy of conferences as we ease out of the summer. One of these events is the annual NLP Summit, taking place October 3-5, virtually. Attracting more than 10,000 of the brightest minds in artificial intelligence (AI), applied natural language processing (NLP), and healthcare, it’s hard to find a reason not to attend the free show. 

Last year, I wrote an article about what to expect at the 2022 event. While some of those themes will carry into this year—large language models (LLMs) and responsible AI to name a few—the conversation around them has completely changed. In contrast to academic research and small-scale projects, this year’s event will focus on applications and best practices learned from putting these models to work in the real world. 

One of the areas experiencing the biggest uptick in AI and NLP is healthcare, which is why there is a full day dedicated to this area. NLP accounts for 39% of the AI in healthcare market share (Zipdo). Why? You’d be hard pressed to find a field in which distilling large volumes of information is more pressing and complex. If you’re already convinced, you can register for the conference here. If you want to learn more, keep reading while I dive into the highlights of the upcoming 2023 NLP Summit. 

Large Language Models in the Real World

It’s hard to read a headline that doesn’t mention generative AI, chatbots, and the language models that fuel them—whether you’re in technology or not. As a quick refresher, an LLM is a type of AI model that is trained on billions of parameters to produce text. Generative AI tools use information from LLMs and other types of AI models to generate new content.

Keynote sessions by The Department of Veteran Affairs and ClosedLoop, respectively, explore applied LLMs in healthcare. The first focuses on a program that applies LLM generative AI interfaces to notoriously messy and inconsistent electronic health records (EHRs). The talk will outline lessons learned and benchmarks used to improve the efficiency and completeness of these workflows, enabling providers to summarize such clinical notes, and then simply ask questions and get quick, explainable, and contextually relevant answers.

The ClosedLoop session will investigate using generative healthcare models to segment high-risk patients into cohorts based on their clinical characteristics and history, enabling providers to better understand risk patterns for individuals. From there, patterns across the broader patient population can emerge, leading to intervention strategies tailored to specific populations. While there’s been tremendous buzz about the promise of generative AI in healthcare, it’s exciting to finally talk about use cases and best practices coming to light. 

Responsibility, Robustness, and Fairness

Responsible AI is once again a highlight of the event program. Much like the Hippocratic Oath, AI should aim to do no harm and create better outcomes for its benefactors—period. This means applying ethical practices before profit, and putting safety, fairness, and robustness at the forefront.

A session by the Ronin Project is a great example of this, highlighting the company’s innovative approach to melding a patient’s unique history, transparency, and the application of causal inference. The outcome is such that each clinical decision is efficient, personalized, and backed by data. This enables empowered healthcare professionals to provide personalized care that enhances outcomes for cancer patients.

Another session from Hyro will introduce the company’s approach to incorporating LLMs into healthcare and beyond. The talk will touch on considerations using LLMs in production, such as privacy, and hallucinations—confident responses from AI that are not rooted in training data. From these lessons learned, Hyro will share best practices for building a stable and reliable system to enhance natural language understanding (NLU) capabilities in the wild.

Open-Source Saves the Day 

The NLP Summit has always given front-and-center stage to the open-source ecosystem, with the first day of the conference fully devoted to such tools. This year is no exception, with sessions covering excellent free solutions for:

  • Vector databases, like Weaviate and Chroma;
  • Responsible AI libraries, like DeepChecks and LangTest;
  • Knowledge Graphs from neo4j and TIB;
  • Low-code and no-code tools, like and the NLP Lab;
  • and, of course, a variety of NLP and LLM libraries and models.

In just its fourth year, the NLP Summit has become the world’s largest gathering of the AI community. As you’ve learned, the program comprises cutting-edge use cases, lessons learned, and challenges of applied NLP having real business impact today. Expert sessions—some mentioned in this article—include top tech and healthcare companies, academic institutions, and AI luminaries. The event is not one you want to miss, so don’t wait—register today.

The Editorial Team at Healthcare Business Today is made up of skilled healthcare writers and experts, led by our managing editor, Daniel Casciato, who has over 25 years of experience in healthcare writing. Since 1998, we have produced compelling and informative content for numerous publications, establishing ourselves as a trusted resource for health and wellness information. We offer readers access to fresh health, medicine, science, and technology developments and the latest in patient news, emphasizing how these developments affect our lives.