Educ(AI)tion: a conversation with Prof Alan Spivey, Vice Provost (Education and Experience)

one of the areas AI has changed quickest, and most drastically, is education. Not only do some students view it as a very patient teacher available 24/7, but more developed models can generate figures, suggest hypotheses and, of course, write reports. In early 2025, Felix spoke to Vice Provost (Education and Experience) and Synthetic Chemistry Professor Alan Spivey on the rise of genAI amongst students, and Imperial’s approach. 18 months and many Claude models later, Prof Spivey reflects on what has changed and the College’s initiatives and policies to educate in an era of AI. 

Teaching, and a level playing field 

As a Professor in Chemistry, Prof Spivey highlights how the field has shifted its view on AI as quickly evolving models became available: “When I spoke previously (18 months ago), we were feeling a little bit smug in Chemistry (...) Now Gemini can take even a badly drawn chemical structure, and it will tell you what it is (...) It's just amazing. The field has really moved on." 

 “I think the capability of AI is clearly evolving very fast, and I think in terms of what that means for the education we provide here, one thing that is really important is that we try and provide as level a playing field as we can for students with different backgrounds and different means. I think it’s incumbent on us to avoid a situation where a student with access to the very latest frontier models gets unfair advantage over a student who doesn't have that", explains Prof Spivey as part of the reason why the College developed commercial system dAIsy, Imperial’s official, secure genAI platform with access to multiple large language models (LLMs) and agents. While he recognises that students can still make use of commercial frontier models, he highlights the College continuously looks for ways to improve on why some students do not use dAIsy. 

Assessments: “we want our students to be able to perform with AI, and in a power cut” 

As opposed to exams, coursework tests students in an environment where the latest tools are available. According to Prof Spivey, “The way coursework in particular is constructed is really important in terms of both how a student is expected to interact with AI to create that coursework and how we assess it.”

He draws attention to the University of Sydney’s “two-lane” approach to assessment, where “AI-off” assessments test that students have themselves developed the required skills for their programme, and “AI-on” assessments focus on engaging productively, responsibly, and authentically with an AI-integrated society. Approving of this model, Prof Spivey explains the College will continue to emphasise the fraction of marks attributed to the coursework elements where assessment relies on a "moment of accountability" (like the Final Year Project vivas) and less so for tasks like report writing. This aims to ensure students focus on understanding and working through complicated processes: “How do you understand it if you haven’t built your way through that kind of pyramid [of knowledge]?” 

Q: How is the university approaching genAI use for staff? Introducing CALT 

A: “It's a great question. I've heard the same comments as well. I don't know if you're aware, but we are just in the process of setting up a Centre for AI in Learning and Teaching (CALT). It's been about a year in the making (...) when I spoke with your colleague 18 months ago, we had just appointed some AI futurists – these were staff individuals in each Faculty – and one of the things that they did was they worked with staff across the organisation [was] to develop a vision for education in the age of AI. I think that's quite an important title, because it's all about education in an environment where AI is readily available." 

The Centre will operate through a human-centered approach, focused on education in an era of AI, currently in process of appointing positions of leadership. The majority of those working for the centre will be Imperial staff from all departments, bought out to work part-time one or two days a week.  

“If we're asking students to be absolutely upfront and clear about how they've used AI, we expect the same of our staff”, he says.  As Prof Spivey explains, Imperial operates through an evidence-based approach, so trials and discussions within CALT will take place to determine what is acceptable for a staff member to use AI for. “There are definitely aspects of feedback that are advantageously delivered by a bot: it's quick, it's immedaite, it's consistent within reason, but it's not human (...) Students are at paying to be at Imperial, they expect that human interaction, (...) but that’s not to say there isn’t a role for some aspects of AI feedback”.

The job market and education: one is changing, so the other must change too 

Over the last six months, Prof Alan Spivey has attended  a steering group co convened between BusinessLDN and Deloitte, comprising leaders from key companies in a wide range of industries. The latest focus of the meetings has been how AI is changing the London workforce. 

The job market is not easy to navigate, and increasingly more skills and AI knowledge is required from applicants: “It is clear that all firms expect the graduates they recruit going forward will still have all the features they have previously (...) But in addition, they will have a deep understanding of AI, technical aspects of AI, they will expect them to understand the inherent biases, environmental challenges, potential for providing hallucinations...".  

When asked about how Imperial’s Career Services will adapt to this changing market, Prof Spivey recognises there is a time-limited period where the College needs to provide students with AI and employability advice over-and-above what is provided within each discipline. 

As part of their goal to provide Imperial students with these required AI skills, the College has recently launched the first part of a 40-hour professional certificate in AI for staff and students, designed to help apply the technology in professional settings and independently develop workflows. 

In the first four days after the launch, 700 staff members (around 8% of total) logged onto the service, with the course made available to students this past Monday.  

About Canvas 

Starting this academic year, Imperial rolled out a Canvas-powered learning management system (LMS), leaving the BlackBoard days behind. Prof Spivey has overseen this transition: “The aspect of Canvas that fits best with the AI era that we're in is that it is the first time we've had a learning management system which is the same whether you're a student studying in the Business School, Faculty of Medicine, Faculty of Natural Science, or Faculty of Engineering. That is great in terms of consistency of experience and also for having a very good experience, because we’ve worked very hard to make sure that Canvas is configured in a way that we hope will prove to be very intuitive". Another advantage Prof Spivey highlights is how  the team is working towards is that modules shared between Departments will now work in a much smoother manner, as he is hopeful about the rate at which we can integrate these features in the system. 

When asked about the biggest setback during this transition, Prof Spivey highlights gathering the requirements of an LMS and going into a procurement process. 

When gathering the requirements, it was not just a matter of cataloguing what BlackBoard and Insendi provided, and “ask[ing] the question ‘what should education in the next 20, 25 years look like?’ (...) There are certain things that we need covering (...), but there's also another set of things that we would really like in a system going forward". 

Let’s keep touching grass

The conversation ended returning to the idea of being human-centric, which Prof Spivey reiterates is “absolutely key.”

“One of the things which I found quite interesting [was] looking at how the widespread use of AI is changing the way we think about the student experience at Imperial", reflectign that it is shedding a light on "non-tech things".

Prof Spivey highlights the people, instructors, coursemates that evolve into lifelong friendships, study spaces and industry opportunities.   "It is more important than ever that we get all those other things right.”