“AI for a Sustainable Future”
The number of stars in the Universe is 10²². The number of possible combinations of dams in the Amazon basin is 10¹⁰⁰⁺. We need AI.
Carla P. Gomes, international AI expert and Feigenbaum Prize winner, joined TU Wien Informatics in June 2023 for two of our most acclaimed public events. The founder of the field Computational Sustainability gave the 2023 Gödel Lecture on June 5 at TU Wien’s Informatikhörsaal, mentored young female computer scientists in an exclusive Meet-up at TU Wien Informatics and held VCLA’s LogicLounge on June 6 together with the esteemed AI researcher Bart Selman.
“AI and Sustainability”
Vienna Gödel Lecture
Fifteen years ago, Carla P. Gomes profoundly changed Computer Science. Against all odds, she won an illustrious National Science Foundation (NSF) grant of ten million dollars, with a vision: to create a new field bridging the gap between computer science and sustainability and revolutionizing how we approach sustainability challenges with technology. Thus, “Computational Sustainability” was created. This emerging discipline harnesses the power of computational techniques, data analysis, formal methods, and AI to address complex environmental and societal issues. From mitigating climate change to optimizing resource management, biodiversity conservation, and AI metabolomics – Computational Sustainability provides a novel framework for scientific discovery and societal decision-making.
Gomes highlighted her transformative research at the Vienna Gödel Lecture 2023 on June 5 at TU Wien’s Informatikhörsaal. Vice-Rector for Research and Innovation Johannes Fröhlich stated that “basic research in AI lays the foundation for groundbreaking advancements. Carla P. Gomes’s exceptional expertise and contributions inspire us to push the boundaries of AI.” Dean of TU Wien Informatics Gerti Kappel emphasized the importance of sustainability, “as a cross-cutting topic influencing all our research, teaching, and daily lives. I am convinced that by embracing sustainability, computer science can lead us toward a resilient future.” Organizer and co-chair of the Vienna Center for Logic and Algorithms (VCLA) Stefan Szeider moderated the talk.
In her lecture, Gomes discussed her current research, above all devoted to fostering scientific discoveries for sustainability. With the sudden advancements of AI in the past year, Gomes is sure that the next frontier in AI will be scientific discovery. For scientific problems, we mostly cannot rely on vast amounts of data used by predominant models like ChatGPT – simply because the data often is not there. This is why Gomes’ focus is on symbolic AI. Machine Learning models rely on data, but symbolic AI works with a so-called “knowledge base”, a collection of symbols and rules that capture domain-specific knowledge. This knowledge base can be created by human experts, express natural laws, or use particular data sets. Symbolic AI models then use logical reasoning derived from their knowledge base to solve problems.
“The great thing about all AI models is that they are transferrable. If you use it to monitor stocks on Wall Street or biodiversity, the technology is the same. We have to use this enormous potential first and foremost to tackle sustainability challenges”, Carla Gomes explains. She uses AI to combat the rapid decline in biodiversity, material science and fertilizers, mapping crops or water supplies, and much more. Her focus is on so-called “deep reasoning networks”, which allow an instant interpretation of data. Gomes is also supported by her colleague SARA, not your typical AI researcher but rather a multi-agent system, which can autonomously generate and tests hypotheses – and the perfect example of how AI can be used for scientific discovery in a multidisciplinary setting. SARA (Scientific Autonomous Reasoning Agent) integrates disciplines such as physics, experimental synthesis, processing, and characterization, along with AI-based algorithms for reasoning and learning. By designing and leading experiments itself, SARA enables rapid materials discovery and development, creating an unprecedented platform for human-machine collaboration.
Gomes’s work serves as a powerful reminder that the responsibility lies with us, not in the algorithms. “I have chosen to improve the world with AI”, Gomes concludes.
“AI and Knowledge”
On June 6, Bart Selman and Carla P. Gomes discussed the connections between Wittgenstein, Gödel, and ChatGPT with Sarah Kriesche from Austrian Broadcasting at VCLA’s LogicLounge. “GPT4 is an advancement that we thought might happen only in another 20-30 years. And in AI research, that usually means we don’t know if it will happen at all,” Bart Selman explains. This extraordinary breakthrough is more than hype because AI models now do not only have an understanding of language but, according to Selman, common sense. Common sense reasoning by large language models has been a dream in AI research ever since the 1960s when John McCarthy envisioned artificial intelligence to make reasonable inferences and navigate the complexities of everyday human life.
But why could this new understanding be such a profound leap? This is where Kurt Gödel comes in. For Gödel, semantics played a crucial role in his exploration of mathematical logic and the foundations of mathematics. His incompleteness theorems, published in the 1930s, addressed the limits of formal systems in capturing all mathematical truths. Gödel’s theorems demonstrated that there are true statements in mathematics for which there is no proof within a particular formal system. That means, within any formal system of logic, there are true statements that cannot be proven within that system – therefore raising questions about the completeness and consistency of any logical or computational model, such as AI. Wittgenstein, on the other hand, focused on the philosophy of language and meaning. In his early work “Tractatus Logico-Philosophicus,” he sought to establish a logical and representational account of language. He proposed that the meaning of propositions lies in their correspondence to the facts of the world. ChatGPT proves that this connection to the real world might not be as important as we thought: “ChatGPT has no real-world approach”, Bart Selman tells us. “Chatbots presumably just use syntactic rules, predicting the next word in a sentence. But still, it ends up understanding the context to some extent, which is remarkable.”
Is ChatGPT a new intelligence? Gomes and Selman think that this is the case. Selman argues that the next frontier will be tackled when ChatGPT starts arguing with itself. “The question will be: how can I let ChatGPT itself verify if its own statements are true? A machine or human has not matched its knowledge and ability to convey this knowledge as of yet – it is the perfect renaissance scholar,” Selman is convinced.
Learn more about Bart Selman and his research here.
“The I in AI”
Young Female Computer Scientists Meet-up
As a female leader in the field and mentor to many, Carla P. Gomes was joined by a group of female PHDs, Post-Docs, and Assistant Professors to have an open discussion about what it means to be a female scientist in a field with challenging numbers of female participation. Here is her advice for all young scientists, especially for women in IT:
- Follow your passion. Pursue things you are interested in, and work hard to make your passion a reality. And remember, good results are 90% persistence and 10% creativity.
- Use your strengths. Know what your strengths are in the first place and take advantage of them.
- Find interesting problems that are at the edge, not common knowledge. This is how you can create a vision and something completely new.
- Generalize. Find a methodology that you can apply to many different problems. Through this, you can advance your methods and become an expert.
- Be positive. Believe in yourself and know that you can do this. Transform something terrible into something extraordinary, and do not give up while doing it, but persevere.
- Network, go to conferences. Approach people and get different viewpoints. This will accelerate your way of thinking and make a massive difference in approaching your work.
- Sell your work. Train how to give talks and write beautiful texts. Confidence comes from repetition, so do not hesitate to take the first step.
About Carla P. Gomes
Carla P. Gomes is an exceptional figure in Artificial Intelligence and Computational Sustainability. She is the Ronald C. and Antonia V. Nielsen Professor of Computing and Information Science, the director of the Institute for Computational Sustainability at Cornell University, and co-director of the Cornell University AI for Science Institute.
With a Ph.D. in computer science from the University of Edinburgh, Gomes has dedicated her career to the pursuit of knowledge in artificial intelligence, particularly in large-scale constraint reasoning, optimization, and machine learning. Her research has had a profound impact on various areas of science. Gomes has become deeply immersed in research on scientific discovery for a sustainable future and, more generally, in research in the new field of Computational Sustainability. For pioneering this new field and her transformative contributions to AI, she was honored with the Feigenbaum Prize by the Association for the Advancement of Artificial Intelligence (AAAI) in 2021. Computational Sustainability aims to develop computational methods to help solve some of the key environmental, economic, and societal challenges to help put us on a path toward a sustainable future. Gomes’s passion lies in leveraging the power of computational methods to address critical environmental, economic, and social challenges.
Gomes’s groundbreaking work has been widely recognized and admired. She has (co-)authored over 200 publications, including influential papers in prestigious journals like Nature and Science. She was the lead PI of two NSF Expeditions in Computing awards. Her research has garnered numerous accolades, including several best paper awards and the ACM/AAAI Allen Newell Award, for contributions to bridging computer science and other disciplines in 2022. Additionally, she was named the “most influential Cornell professor” by a Merrill Presidential Scholar (2020). Gomes is a Fellow of the Association for the Advancement of Artificial Intelligence (AAAI), a Fellow of the Association for Computing Machinery (ACM), and a Fellow of the American Association for the Advancement of Science (AAAS).
About Vienna Gödel Lectures
Named after the famous Austrian-American logician, mathematician, and philosopher Kurt Gödel (1906-1978) and introduced in 2013, the annual Vienna Gödel Lectures bring world-class scientists to Vienna. The lecture series illustrates computer science’s fundamental and disruptive contribution to our information society, and it investigates how our discipline explains and shapes the world we live in—and, thereby, our lives as such.
The series of public lectures LogicLounge continues to bring together the general public and experts from the fields of logic, philosophy, mathematics, computer science, and artificial intelligence. Since its inception at the Vienna Summer of Logic in 2014 – the largest event in the history of logic – the series has since been traveling between Vienna and the venue of the CAV (International Conference on Computer-Aided Verification), where it has already become a regular event in honor of Helmut Veith (1971-2016).