The next in the series of SCI’s free online talks looks at what chemists need to know about AI.
Dr Zoë Webster will discuss the meaning behind some of the key terms in the AI space and how they apply in real-world scenarios - and what matters most for successful and responsible AI use.
C&I magazine spoke to Dr Webster, who advises organisations on AI strategy, practice and governance, having been in the AI space for over two decades as a practitioner, leader and researcher. She previously built and led BT's AI Centre of Enablement, delivering AI from concept to operation across numerous use cases. Before that, she directed national AI and innovation strategy at Innovate UK.
C&I: What is the real impact of AI right now and where do you see it going next?
Webster: This is the big question – what impact is AI really having? AI, in its broadest sense, is not new and has already made it possible for organisations to engage with the right people, in the right way and at the right time. It allows us to search for information from an ocean of data and content. And to identify anomalies in systems, networks and processes that warrant further attention.
For Generative AI specifically, understanding its impact is currently proving harder to gauge that it should be. In many organisations, those adopting AI are reporting time savings, for example, as they use LLM-based tools such as ChatGPT, Claude and Gemini to generate first drafts, edit copy and to summarise meeting transcriptions, but that isn’t always being translated to impacts on the bottom line.
In science, as in many other domains, AI can help with the language-heavy tasks, such as summarisation and drafting, and it is easier than ever to instruct or query machines. However, AI is not, yet, so widespread and embedded in running and managing the more physical processes involved in chemistry including lab execution, data collection in the field and clinical validation.
AI can be helpful in generating new ideas and paths to test, and makes it easier to write papers, but there is then a bottleneck downstream, where the rubber hits the road. More advances in Physical AI, AI that can work with and/or within physical systems, may start to power more ‘Self-Driving Labs’, laboratories that can operate somewhat autonomously to run experiments more quickly.
To do this, AI may need to learn from the scientists who do the physical stuff. This would enable AI to drive the discovery and validation cycle faster, limited mainly by the time needed for any chemical reactions. LLM-based applications may help in enabling scientists to instruct and manage such systems.
What do scientists need to do to harness and benefit from AI?
AI is a broad field and while a lot of attention is currently on Generative AI, such as LLMs, and Agentic AI, there is more to it than that. To get the most from AI, it is always useful to understand and break down the opportunity. What is it that you are trying to do?
Scientists should hone their skills at understanding the intermediate steps of problem-solving so they can map the right AI tool to the right job.
Generative AI is incredibly powerful and particularly useful for translating, summarisation and drafting. Where repeatability and accuracy are the main goals, or where you need to account for hard physical constraints, then Generative AI may not be the best tool and other approaches are available. This goes further than ‘prompt engineering’ and into process mining and engineering. This is something that may be natural for many in science and a key skill for harnessing AI effectively and sustainably.
How do you see this evolving over the next few years?
I am excited about neuro-symbolic approaches in AI that could combine the power of Generative AI with other, more logical and rule-based forms of AI, as this could help with greater scientific validity and explainability.
Also, as I mentioned earlier, developments in Physical AI may help to close the loop between prediction and the lab bench through autonomous experimentation and that is where we might start to see real and scalable impacts from AI-powered scientific discovery.
Human insight and perspectives will remain vital and we may start to see more regulation around agentic AI and foundation models (such as LLMs) including on transparency, guardrails and environmental impacts.
This SCItalks session: What chemists should know about AI and where it may be going takes place online on Wednesday 23 September at 4pm, and is free to attend.
To find out more and register for this event visit the SCItalks page on the SCI website.
SCI’s SCItalks are given by the most prominent and influential academics, thinkers and industry leaders who are working to translate scientific innovation into societal impact. These talks form part of SCI’s charitable outreach which seeks to create a public conversation about science-based topics in a way that is accessible to all.
Further reading on AI and chemistry
- Creating chemistry that machines can understand
- Ten scientific breakthroughs getting ready to change the world
- Quantum technology is reaching a commercial tipping point for chemicals and pharma
- How woolly mammoths, ancient penguins and AI are helping the hunt for new antibiotics
- AI in pharma: Helping science to find drugs faster
Chemistry & Industry (C&I) magazine reports on the people, the scientific advances and the industrial innovations being harnessed to tackle society's biggest challenges. C&I covers advances in agrifood, energy, health and wellbeing, materials, sustainability and environment, as well as science careers, policy and broader innovation issues. C&I’s readers are scientific researchers, business leaders, policy makers and entrepreneurs who harness science to spark innovation.
Get the latest science and innovation news every month with a subscription to Chemistry & Industry magazine. You can subscribe to C&I here.
Visit Blog