All programmes Programme · 28–30 August 2026

AI in Practice

Three evenings for people with a thesis to finish — and they stayed past the end of every one.

176signed up in fourteen days, from one announcement
3 of 3evenings ran past their scheduled finish
6 in 10were still in the room when the clock said stop
01The three evenings 19:00–21:00 WAT each night. Free, online, and recorded for replay.
Friday — where the hours go

Where AI actually is right now, and which parts of a research week it can take off you.

Saturday — the research pipeline

Literature and referencing, then data and analysis. End to end, on the free tier of everything.

Sunday — producing the work

Slides and documents that do not look machine-made, and building the small tool your department keeps asking for.

02How it went Counted from the per-session join and leave times, notetakers excluded.

They stayed to the end. Every night.

People in the room at each five-minute mark, 19:00–21:30 WAT

Three nights, three different room sizes — and the same shape every time. The room fills by half past seven, holds flat for two hours, and is still full when the clock says stop. Nobody drifts out at half time.

Still there when time was up

Share of that night’s peak room present at 21:00

Sixty-four, sixty-two, fifty-eight. The retention barely moved across the three evenings — a smaller room on Sunday behaved exactly like a bigger one on Friday.

03Who signed up From the registration form, 176 responses.

Who was in the room

Registrants by discipline — health sciences, 78% of the cohort

Not a general audience. Nearly eight in ten were health-science students or clinicians, and seven in ten came from one university.

What they came to get done

Free-text answers, coded

The top three are one job: finish the thesis. Ten people wrote some version of “stop paying someone to do my analysis”.

How fast it filled

Registrations per day, 14–27 August

81 people in the first 48 hours — 46% of the cohort — from a single announcement and a WhatsApp thread. No advertising, no budget.

04Why the lab runs it
Training is how we meet the people we build for.176 health-science students told us, in their own words, which parts of a research week hurt. That list is the backlog for Coursified and Hypatia. A cohort is cheaper than a survey and more honest than one — people tell you what they actually want when they have given up three evenings to get it.

Run this for your department or your team.

Built around the work your people actually have to produce. The next cohort is not booked, which means the date is yours to pick.