I love exploring technical topics by writing about them. But, a lot of data science work involves explaining the breadth of your complicated model in 20 minutes at a meeting. So, I set a goal of becoming a better public speaker.
I’ve given a lot of conference talks, but this has been my first year being on both podcasts and public radio(!). What’s exciting and hard about audio (especially when it’s live) is that you have to express yourself clearly in a limited amount of time, knowing that the recording will then be around for posterity. And man, do I hate listening to the sound of my own voice.
As a side note, a resource that really helped me prepare for these was Paul Ford’s Guide for People Doing Media Things.
I’ve recently been on three different podcasts/radio shows, and hopefully it’s not too obvious which one was first and which was the most recent ;).
- Adversarial Learning - We talked about data science myths, and went on a lot of fun tangents. It’s a lot like having a drink with a friend in industry and just shooting the breeze, but you end up learning something, too.
- Techtonic - This one was live radio, and Mark and I discussed the mechanics of sharing your day-to-day life on Facebook, what kinds of algorithms Facebook uses to keep track of you, and the general implications of data collection. And we had a call-in all the way from Norway!
- Data Lab - I talked about my recent Alice in Python packageland post, what T-shaped skills are, and what I enjoy about consulting.