Last week, Google hosted its annual I/O developer conference, showcasing the company’s vision for its products and services. Usually this involves exciting conversations about Google’s mobile platform, Android, or the latest hardware the company plans to release. However, something else stole the show this year: Google’s incredible advances in artificial intelligence (AI).

AI has been gaining traction in Silicon Valley and around the world in recent months, especially with the release of Chat-GPT and advanced generative AI systems. This has set off a somewhat frenzied rat race among tech giants: everyone wants a piece of the AI ​​pie, and the race to the top is intense.

For example, Microsoft is investing significantly in its AI capabilitiesrelentlessly trying to integrate technology into various sectors, from manufacturing to healthcare. Amazon is also innovating in this space, offering clients turnkey services to build their own machine learning and AI models.

Similarly, Google has made incredible strides in its own AI journey. While many may suspect that the buzz around AI is relatively new, the company has actually been laying the groundwork for advanced language models and usable AI for many years. This work has culminated in products with incredible potential, including PaLM 2Google’s latest generation big language model.

As an introduction to PaLM 2, Zoubin Ghahramani, Vice President of Google DeepMind, wrote last week: “today we’re introducing PaLM 2, our next-generation language model. PaLM 2 is a state-of-the-art language model with improved multilingual, inference and coding capabilities.” He further describes these capabilities, discussing how PaLM 2 is trained in 100 languages, has a greater ability for common sense reasoning and advanced mathematics and logic, and can even generate advanced code.

Notably, Ghahramani also discusses one of the most anticipated arenas of AI impact: healthcare. It enters Med-PaLM 2, a large language model developed specifically for generating medical insights: “trained by our medically knowledgeable healthcare research teams, [Med-PaLM 2] can answer questions and summarize insights from a variety of dense medical texts. He achieves top scores in medical competence and was the first major language model to perform at the “proficient” level on the American-style medical exam. [USMLE] questions. We are now adding multimodal capabilities to synthesize information such as X-rays and mammograms to one day improve patient outcomes. Med-PaLM 2 will be opened up to a small group of Cloud users for feedback later this summer to identify safe, useful use cases.”

You can watch the complete opening speech for 2023 here:

Google Cloud, especially for healthcare, has been incredibly successful over the past few years. For example, Google Cloud Healthcare Data Engine is an industry-leading platform for optimizing healthcare interoperability, which could solve some of the sector’s most challenging problems: significant data fragmentation, lack of cohesive insights, and fragmented patient journeys.

Google Cloud Global Director of Health Strategy and Solutions, Aashima Gupta, and Global Director of Health Plan Strategy and Solutions, Amy Waldron, explain: “Industry-friendly LLMs like Med-PaLM 2 are part of a growing family of generative artificial intelligence technologies that have the potential to significantly improve healthcare experiences. We look forward to working with our customers to understand how Med-PaLM 2 can be used to facilitate rich, informative discussions, answer complex medical questions, and find insights in complicated and unstructured medical texts. They could also explore its utility to help craft short and long answers and summarize documentation and insights from internal datasets and bodies of scientific knowledge.”

The cloud team was also responsible for the launch Medical Imaging Suitea comprehensive AI platform that will make medical image data “accessible, interoperable and useful”.

Without a doubt, the development of PaLM 2 will change the game in healthcare. Whether it is used to perform advanced analysis on vast amounts of public health or patient data, or simply as a tool to better triage and improve physician workflows, technology has enormous potential to create tangible impact.

Especially in terms of synthesizing and generating insights, this technology can be extremely useful for organizations looking to increase their tactical approach to improving patient outcomes. For example, for public bodies, the use of this technology may one day enable the ingestion of terabytes of otherwise unstructured public health data to generate useful insights. For private organizations, this technology could prove to be a significant boon in the areas of interoperability, improving the patient journey, and succeeding in truly longitudinal care.

However, this technology requires guardrails. Although there is considerable potential for beneficial use, there is also some “fear of the unknown”. That is, in the wrong hands, the powerful technology driving these advances in AI can have negative consequences. This is the main reason why the developer conference had an entire segment dedicated to “responsible development” of AI, serving as a reminder of the checks and balances needed in this rapidly growing arena of technology. However, if developers and innovators are able to create and use this technology in a safe and sustainable way, it could potentially change the face of healthcare delivery for generations to come.