New way to explore the brain in all respects complexity and in one summary until now impossiblesystem that is able to help scientists to formulate a new one hypothesis and evaluation in connection with thousands of data and articles, tools that help doctors read X-ray and CT images more precisely: these are some of the examples presented today in Warsaw at the Google Research@Poland conference, the only scientific conference organized this year by Google in Europe.
At the ancient Norblin factory where silver, brass and copper were worked in the 19th century, which has now been discovered and is a gem of industrial archaeology, researchers from around the world meet to explore the future of science in the age of artificial intelligence. The sensation isopportunity From “open A new age at what time talents are needed“, said Magda Kotlarczyk, country manager of Google Poland, when opening the conference.”It’s time to invest in research For produce A new growthhe added.
The purpose is develop models That integrate more tools and that they are capable of doing so tackle complex experiments. And an example of the advantage of complexity is brain researchorgans that have a the number of cells is 20 times greater than the population on Earthsaid Lizzie Dorfman, Google Research Group Product Manager.
Meanwhile, researchers from any scientific discipline can experiment Fellow AI scientistthe new one model from AI designed for laboratoriessaid Google researcher Annalisa Pawlosky. “This speeds up the clock of discovery because its important innovation is to collaborate with researchers, contrary to what the automated systems tested so far have done”. Among other innovations, these systems “take into account cognitive aspects, propose hypotheses and interact with scientists, in some cases suggesting how to overcome certain obstacles”.
However, there are limitationshe added, because system is trained with scientific literature and, most importantly, it has a limited access to negative resultfar fewer were published. For Tiago Dias da Costa, from Imperial College London, the ability to rely on the help of effective tools can encourage researchers to publish even negative results. “By formulating the right hypothesis – according to his observations – a researcher can save 90% of his time, there will be more resources to do other things and research will be faster”.
Meanwhile, he also debuted MedGemmathe latest product among AI models capable of reading images for medical diagnosis, from x-rays to mammograms.
Reproduction protected by law © Copyright ANSA
