Friday 7 November 2025, The Alex Hotel, Dublin City, ConnectED 25 brought clients, partners, and friends of Cobblestone together for an afternoon of lively discussion and hands-on exploration of what great learning looks like today. From our bingo-style icebreaker to fireside chats and a closing panel on innovation, the energy in the room was all about practical ideas you can use right away.
Below is a short recap of our learning industry event, the key takeaways we heard across sessions, and useful next steps if you would like the resources or to continue the conversation.
The event at a glance
- Welcome lunch and registration set the tone for a relaxed, collaborative afternoon.
- Icebreaker helped everyone connect quickly and surface shared challenges and wins.
- Panel: Many Voices in Learning explored how to hear, balance, and act on the diverse perspectives in any learning ecosystem.
- Fireside chat: AI in eLearning looked at practical use cases that add real value for learners and teams.
- Fireside chat: Branding, localisation, and realistic scenarios showed how design and context lift relevance and performance.
- Lightning talk: Implementing AI in your organisation offered a compact roadmap for getting started responsibly and how L&D can lead the way.
- Panel: Innovation in learning unpacked what innovation means in practice, how to repeat it, and how to measure impact.
(Speakers included leaders from The Department of Agriculture, Food, and the Marine, Learnovate, Adult Literacy for Life, Krisolis, The Institute of Public Administration, and the Cobblestone team. Thank you all for sharing your time and expertise.)
Five things we learned together
- Start with voices, not tools
The most effective solutions surfaced the needs of different learner groups, managers, and stakeholders before choosing formats or technology. Listening frames the problem and prevents beautiful solutions to the wrong problem. - Make AI small, safe, and useful
The most convincing AI examples were narrow in scope, transparent in use, and designed to support people rather than replace them. Think content drafting with human review, structured feedback for practice activities, and smart routing for learner support. - Design for cognitive ease
Branding is not decoration. Consistent visual systems, plain language, and purposeful interaction patterns reduce cognitive load and increase completion and transfer. - Localise beyond translation
Localisation means voice, scenarios, regulation, and culture. The best results adapt tone and example sets, not just words, so learners recognise their world in the learning. - Treat innovation as a process
Innovation is repeatable when it has constraints, governance, and measurement. Pilots, clear hypotheses, and agreed success metrics are what make “new” become “better”.
Take a look at some session highlights
Bingo-inspired icebreaker
We opened the learning industry event with a fast networking game that paired people around prompts such as “has fought learner fatigue and won” and “works in a completely different industry than you” Within ten minutes most attendees had spoken to five or more new contacts. The activity surfaced themes that reappeared throughout the afternoon.
The discussion focused on balancing the needs of learners, managers, compliance owners, and technology teams without diluting outcomes. Practical tactics included short discovery interviews with representative users, lean prototypes that test the riskiest assumptions first, and decision logs that record trade-offs. The group highlighted the value of transparent governance so stakeholders see how input is weighed and what will happen next.
AI adoption franed through the lens of L&D as the catalyst, not waiting for a perfect enterprise blueprint, but setting the standards, running the pilots, and proving the value. L&D owns the playbook, quality bar, and change comms, partners with IT/Legal for data and risk, and scales only where results warrant it.
The closing panel treated innovation as a repeatable process rather than a one-off. Panellists compared models for turning insights into backlogs, prioritising experiments, and measuring impact beyond completion rates. Useful metrics mentioned included time to competence, first-time-right performance on critical tasks, and help-desk ticket trends after release.
Want to learn more?
If you would be interested to find out more about the day; the insights and demos that were shared at the learning industry event, we would be happy to discuss on a quick call. Fill out a form on the contact us page here.