Earlier this year I got excited about an AI release (https://thekoalanews.com/ive-seen-the-future-and-its-not-vibe-coding/), where the technology moved from chat bots to agents – albeit still requiring a fair bit of technical know how to get anything from them.
For six months I’ve used Claude like an assistant, coded some tools, stopped using Google for search, but never really got to the point where I could set an agent, or team of agents, on a task to achieve a goal. That’s not to say it couldn’t be done – it just required more technical ability than I have.
In the last month we have seen another pivotal moment in how AI will develop with the release of more ‘consumer’ friendly personal assistants. Products like GrokBot, Muse, Dots, Instinct and Hermes will now happily perform whatever tasks need to be done to achieve a set goal. This includes running up multiple agents to help perform the task and managing them until the goal is achieved.
Why is this important?
Well, for the first time, I think we can see a future where we talk to our agentic personal assistant to help us get things done. It’s still very early days. GrokBot is only available on the most expensive plan, and many might not want their personal life run by a company associated with Elon. And Muse, from Meta, has only just been released for the US and people might also be concerned about putting their personal life in the hands of Mr Zuckerberg.
Even though there are still many hurdles to overcome before wide-scale adoption, it’s hard to rule out that this is going to be how we use technology in the future. So, we need to start thinking now about what this means for how we manage international recruitment and admissions. Because it won’t be far away before someone talks to their phone, glasses, home console (Apple just released their HomeView product this week) and says, ‘based on what you know about my experience and what my goals, find me a program of study that helps me achieve my goals, fits in with my lifestyle, budget and commitments and enrol me on the selected programs.’
Up until now we’ve been dealing with generative or genAI: how we use chatbots and GPT’s to generate a response to a question or automate a task. What’s becoming possible is true agenticAI. Tasking a digital assistant to complete the specified task. Without needing to know how to set up a virtual machine.
And it raises a lot of questions.
How do recruitment and admissions teams respond to an AI agent enquiry? Do we reject it entirely and say we only deal with humans?
What information do we make visible to this AI agent and is it current, or is it based on a landing page set up by a faculty that is now out of date? Does it answer the questions in a way a digital assistant wants to know, or does it contain too much marketing fluff, which GPT’s will discount?
What workflow do we put the request through, what documents and information will we need, how will we collect it, what sources will we trust and who will make the ultimate decision?
And if we can’t offer the exact program, can we put a program together with a partner provider that delivers the exact requirements of the potential learner? Can we then manage the administration with that partner automatically – including credits, recognised prior learning, billing and assessment?
This might seem 10 years away, but so much has changed since ChatGPT launched four years. 10 years could easily become five.
Are you ready for that?
A lot of the questions (and problems) we imagine might deliver more immediate benefits. That justifies spending some time thinking about them, starting to prepare, and putting us in the best possible position to take advantage of the agentic future.
For example, working with marketing and recruitment to ensure a single source of truth for program information. This makes a university more ‘AI visible,’ which helps now, but also sets up good processes and governance for ensuring the right data will be accessible by that agentic personal assistant in the future.
Recently, I asked my GPT to help me research possible future study ideas. Some options really excited me. And then I clicked on the sources. One course only ran once in 2017, the others all showed information about programs in 2026 whose deadline had already passed. They gave me no insight into when the course might run again. I had to dig really deep into the provider sites to get the information I needed or fill out a generic ‘interested in this course’ form, which signed me up for general updates about the university and not the program I was interested in.
I didn’t want that. What I did want was more specific answers to my questions – like start dates, costs, units and ideally recognition of prior learning, or credits so I’d know the actual length and cost of the program for me.
This is an opportunity lost. It’s also a good illustration of why we need to also rethink why we do what we do, and how we can leverage the emerging technology to create a competitive advantage. We can’t just automate what we are doing now. Even though that does deliver important time saving benefits and is worth doing.
New generative AI tools, like Jev, will make automation and developing workflows easier for internal teams. Which means we can start to ask ‘what if,’ and run micro-experiments. For example, you could run a trial to see the impact on conversion of providing highly personalised responses to an offer, based on the information already known and collected about the applicant. This starts to get an admissions team thinking about the automated customer experience, even if the customer is an agent.
I was amused to see Ethan Mollick set up a separate page for agents to promote his upcoming book on working with agents (https://co-existence.ai/for-ai). If our customers are a computer – if the agent is empowered to make a decision on behalf of its human – how does that change our workflow?
And while we’ve been pretty good at using tools to automate repetitive and manual steps in the admissions process, it’s also important to start thinking about the purpose of recruitment and admissions for each institution, and the key metrics they need to be measured on. This can be used to help create a process, with the assistance of AI, that delivers on that goal. This moves us towards thinking about how admissions can be transformed with technology to produce a better solution for all, rather than just sped up. And how we design the process so a named person still controls and is responsible for the outcome, just as if the entire workflow was physically done by a person.
Sometimes it’s hard to find the time to lift your focus from the day-to-day battles to think about what’s ahead. Yet, and especially given how long higher education transformation projects can take, that 10 year version of the future will be on us before we know it.
That’s why we’ve put together a panel to discuss some of these topics at AIEC: An idealist, a realist and a futurist walk into a bar in 2030 – to talk admissions. I’m intrigued to see where that conversation will take us. https://www.xcdsystem.com/aiec/program/sV0TkNj/index.cfm?pgid=5906&sid=55653
Jason Howard founded StudyLink, the world’s leading admissions and agent network platform, in 1992. He steered that business through three profound technology transitions – from print marketing to digital marketing, on-premise to cloud hosted software, and licence fee to SaaS business model – before successfully exiting in 2024. Having managed customers and users through these transitions over 30 years, Jason is in a good position to consult and advise on how the higher education sector navigates the next wave of technology change.
