UCI IT Department Organizes Seminar on Artificial Intelligence Systems with Dr Yolanda Gil | New university
The UCI Computer Science Department hosted Dr Yolanda Gil, Director of New Initiatives in Artificial Intelligence (AI) and Data Science at USC’s Viterbi School of Engineering, for a Computer Intelligence Seminar artificial on November 12. Gil discussed the future role that artificial intelligence systems may have as authors of scientific papers to produce their own experiments and to shape interactions between scientists and publications.
Gil opened the discussion by introducing the topic of artificial intelligence and the ways it has been used over the years.
âAI has made many important contributions in the intelligence, modeling, capture and representation of knowledge itself,â said Gil. “In the area of ââknowledge representation and ontology, for example, AI has been able to connect and integrate so much knowledge in biology across the genome.”
According to Gil, the beneficial use of AI was prioritized over a scientist’s ability to conduct scientific work. Inefficiencies in approaching problems from a human point of view allow errors and biases and lead to poor communication of scientific concepts.
âWe think of scientists as very special thinking people. As humans however, I see that we sometimes suffer in this area. Humans are not very systematic because we are not thorough when performing a task, âGil said. âWe also make mistakes and mistakes. This causes the articles to be written in a way that others cannot rely on. Our way of doing science is essentially non-ideal. AI, in this way, can help us become better scientists. “
For AI systems to eventually become scientists themselves, article generation should be a task that they can consistently perform from a benchmark or by themselves. Highlight some of the ways scientists conduct research, collect data, and write scientific papers. Gil explained these concepts to discuss the possibility of detailed and accurate AI reporting.
âSpecial software can be used to perform experiments. Once the data is obtained, the results are sorted and combined to eliminate anything that is insignificant. This way a report can be written. Highlighting the right principles for archiving and sharing data and workflows with AI systems are all important parts of producing a document, which humans don’t generate, âsaid Gil.
Gil went on to highlight the ways AI systems began to shape and develop scientific papers. The first example she described was scientific vocabulary and the ways it can be used to describe scientific data in the field of paleoscience, the study of climatic and environmental processes.
âWhen you try to describe the climate trends over thousands of years, you want to put all the data together to report the results. To unify their data, paleoscientists seek to speed up the process with AI. We offered them a new approach where we could crowdsourcing the way each community of scientists wanted to describe their data, âGil said.
In this procedure, a face to face meeting was held where all professionals agreed to use the system to help write the report. A few months later, a common way of describing their data was developed. Gil explained this system.
âA systematic wiki was developed, where these scientists could adopt terms that existed. They could also browse and add their own terms, âGil said. âMany standard ways of describing factors like location, for example, have also emerged. This approach to AI has had to adapt to changes in terms along the way. “
âIn regular scientific articles, we are able to describe the general method used in relation to its execution in a study. We use rules and patterns in technologies and languages ââthat impose constraints on the workflow, allowing users to be consistent with the data they present, âsaid Gil.
For example, time series analysis, in which an analysis of a sequence of data points over a period of time is discussed, can use a workflow with AI systems.
âWe represent all the steps as an abstract model that provides specific steps for the constraints that accompany them with AI. These algorithms developed for constraints are good instruments for obtaining the knowledge to build these workflows, which can be applied in many contexts, âsaid Gil.
To conclude, Gil explained that we will have to deal with complex scientific phenomena in the future. According to Gil, scientists and AI systems will need to learn to become discovery partners in the scientific community.
âAI systems can’t just be about being told by scientists what to do. I think AI systems need to learn on their own, learn more, notice new software, and connect with people or other AI systems. Reproducing and writing articles is a very modest goal, âsaid Gil.
Korintia Espinoza is a STEM intern for the fall term of 2021. She can be reached at [email protected]