Google's Latest Experiment: AI Chatbot Being Tested To Answer Medical Questions
Sundar Pichai , CEO of Google , has announced the development of PaLM 2, a large language model (LLM) with improved multilingual proficiency, enhanced reasoning abilities, and coding capabilities. This AI model, trained on 100 languages, is designed to perform a wide range of tasks. During the announcement, Pichai emphasised that PaLM 2 would be integrated into Workspace apps, Moed-PaLM for medical purposes, and Sec-PaLM for security applications. Excitingly, recent reports indicate that Google has been testing an AI chatbot specifically designed to expertly answer medical questions.
According to sources cited in a report by The Verge (via The Wall Street Journal ), Moed-PaLM 2 has been undergoing testing since April at renowned research hospitals such as the Mayo Clinic , a non-profit organisation based in the US. The goal is to assess the model's potential in aiding areas with limited access to doctors. Greg Corrado , Google's senior research director, acknowledged that Moed-PaLM 2 is still in its early stages.
Corrado expressed caution, stating that he would not personally rely on Moed-PaLM 2 for his own family's healthcare journey. However, he firmly believes that the model has the potential to exponentially expand the application of AI in healthcare, where its benefits can be truly impactful.
Google asserts that Moed-PaLM 2 outperforms general chatbots like Bard, Bing, and ChatGPT in healthcare conversations. This superiority is attributed to Moed-PaLM's training on a carefully curated dataset of medical expert demonstrations, ensuring a more specialised and accurate understanding of healthcare-related queries.
The report also highlights that customers participating in the Moed-PaLM 2 testing phase will have control over their data. Google guarantees encryption and affirms that it will not have access to the collected data. Data plays a crucial role in training AI chatbot models, and earlier this week, it was reported that Google updated its policy, explicitly stating that publicly available information from the open web will be used to train Bard and other AI models. Google confirmed this update, with spokesperson Christa Muldoon emphasising the long-standing transparency of their privacy policy in utilising publicly available data for language model training services such as Google Translate .
With the introduction of Moed-PaLM 2, Google aims to advance the capabilities of AI in the healthcare domain, promising enhanced language processing and specialised medical expertise. As the testing and development of this innovative AI chatbot progress, the potential for improved access to reliable medical information and support in areas with limited healthcare resources appears brighter than ever before.
According to sources cited in a report by The Verge (via The Wall Street Journal ), Moed-PaLM 2 has been undergoing testing since April at renowned research hospitals such as the Mayo Clinic , a non-profit organisation based in the US. The goal is to assess the model's potential in aiding areas with limited access to doctors. Greg Corrado , Google's senior research director, acknowledged that Moed-PaLM 2 is still in its early stages.
Corrado expressed caution, stating that he would not personally rely on Moed-PaLM 2 for his own family's healthcare journey. However, he firmly believes that the model has the potential to exponentially expand the application of AI in healthcare, where its benefits can be truly impactful.
Google asserts that Moed-PaLM 2 outperforms general chatbots like Bard, Bing, and ChatGPT in healthcare conversations. This superiority is attributed to Moed-PaLM's training on a carefully curated dataset of medical expert demonstrations, ensuring a more specialised and accurate understanding of healthcare-related queries.
The report also highlights that customers participating in the Moed-PaLM 2 testing phase will have control over their data. Google guarantees encryption and affirms that it will not have access to the collected data. Data plays a crucial role in training AI chatbot models, and earlier this week, it was reported that Google updated its policy, explicitly stating that publicly available information from the open web will be used to train Bard and other AI models. Google confirmed this update, with spokesperson Christa Muldoon emphasising the long-standing transparency of their privacy policy in utilising publicly available data for language model training services such as Google Translate .
With the introduction of Moed-PaLM 2, Google aims to advance the capabilities of AI in the healthcare domain, promising enhanced language processing and specialised medical expertise. As the testing and development of this innovative AI chatbot progress, the potential for improved access to reliable medical information and support in areas with limited healthcare resources appears brighter than ever before.
Next Story