There is much emerging evidence that clean, consistent data is the key to the good performance of LLMs. Many parties, such as the IEEE, HippoAI, and GHDIF, have reached out to us at openEHR as they understand the importance of our work for the future. AI is only as good as the data it receives and is trained upon – so great data reduces the risk of hallucinations and false results. Open concepts also extend to the training of AI, such as having standardised ways to describe the data sets and demographics used, all of which are enhanced with openEHR. We are also seeing the emergence of more Open AI models on the market (and no, Chat GPT is not an open-source model, despite the company being called OpenAI), as they are more transparent and supported by communities like ours. These provide more assurance to health systems and transparency in terms of how they are constructed.
Aside from AI, we are seeing a huge growth in geographical interest, with openEHR affiliates being set up across the globe. I have been assisting several new affiliates across the world, all of whom want to create better health data platforms for their citizens. The mistakes I and others have made in the past, which have led to data silos, are now being recognised, and countries are now valuing their data as a civic asset. Life sciences and research are also joining conversations, from making trials easier to finding rare disease patients, so there are many benefits to our approach.
I am really looking forward to working with this community to build the future infostructure for health and care. This will be the infrastructure that enables the future of healthcare systems.
Rachel Dunscombe, CEO, openEHR International
Guest editorial featured in the Mission magazine, volume 1.