This months AI^2 forum was about how to ‘Communicate your Work to a Non-Technical Audience’…

This months forum was highly motivated by a lot of talk whirling around our PhD environment about how:

a) Our models are wrong or not good enough for implementation in the real world

b) End-users don’t understand it

c) Academics fail to engage with the public and make an impact

Which are all seeming very damning!

To answer this we looked at our own work. And how typically we focus predominantly on the centre of the academic papers, i.e. the method and we get lost in the technical weeds. But actually we found that focussing on the ‘edges’ of the paper is what is most important when engaging with the real world, i.e. the conclusion and initial motivation.

After briefly going over this whirling indictment of our work, we spent the vast majority of the two hours in a highly interactive manner. Before the tasty food arrived (not pizza this time but perhaps devoured even faster?) we had a lively time creating a single powerpoint slide poster of our work, as if we were on presenting to a non-technical audience and focussing on the conclusion, challenges to implementation, and problem motivation. The fun bit was that we could only use generative AI to make the posters! Thus seeing how far generative AI has come, and collectively finding new Generative AI tools to help each other make an engaging and accurate poster.

After the food (woo!) we played a game of Broken Telephone, where we would explain our projects to one other person, and then listen to theirs. Then we would relay their project on to a new person and hear the pitch they were also relaying… and so forth, very confusing and quickly much information was lost through the pitches!

This game of Broken Telephone was useful in seeing how memorable and communicable your project pitch was, and we could really see how easily people absorbed and took in what we were saying, as they tried to repeat it to someone else!

Finally, we planned to write down the corrupted project pitches and use more generative AI in the form of AI voices to present the pitches next to their respective posters! Some people chose highly peculiar and interesting voices to read out the pitches after they had been passed through the broken telephone.

We were running out of time, so we got the humans who heard the pitch last to present in front of the respective posters which was interesting to see what information was retained.

Overall it was great fun, and really made us realise how confusing our work is and hard to explain. Hopefully we can go forwards and better communicate our work to a non-technical audience!

‹ ›

Download PowerPoint Presentation

Blog written by: Sam Llanwarne