Last week, we received the session and speaker ratings for the GDC AI Summit. We wanted to share some of the highlights with you.
First off, of the 9 summits this year, the AI Summit was the 2nd highest-rated one. Our average score was only 0.02 behind the Indie Summit (and they’re a hard act to beat!). This year’s scores were also better than our scores from last year. (Note that despite the fact that there were only 4 choices on the form, GDC put the scores on a 5-point scale to align them with the historical numbers.)
Additionally, we are really growing into the bigger room that they started giving us. Our average attendance was 272 this year — although that might be slightly lower than it should be because of some logistical problems with counting. Compare that to 238 last year and 203 the year before and you can see that this is really taking off! Our top session brought in a whopping 392 people!
Once again, most of our sessions were very tightly clustered with near-identical average ratings. However, we like pointing out some of the top ones that stuck out:
Once again, this year, the rant session was the highest-rated session with a rating of 4.73 (out of 5). (In the “you can’t please all the people all the time” category, 2 people rated the session as “poor” – presumably one of which is whoever left “boring” as their comment.)
The 2nd highest session (4.63) was the tactics session by Alex Champandard, Mathew Jack, and Philip Dunstan. While their content was admirable, the comments also suggest that people were mesmerized by their incredible presentation method!
With a 4.56, the #3 session of the year was the “off the beaten path” session with Chris Jurney talking about computer vision, Mike Robbins on neural nets in SupCom2, and Ben Sunshine-Hill on using AI to determine the proper level of on-screen detail.
As I mentioned, a good number of the sessions were packed in with similar scores right behind the 3rd place one. So close that they’re almost identical. Congrats to all the presenters across the board… good stuff, folks!
Because we have so many people that were in multiple sessions, and so many sessions with multiple people, there isn’t always a direct mapping of session scores to presenter scores. To solve that, I look at who had the highest per session score as well as who had the highest aggregate score across all their appearances. Steve and I would like to recognize the highest people in those lists.
Highest Individual Lecture Performances
For the 2nd year in a row, Brian Schwab’s rant knocked it out of the park. His 4.65 rating was the top individual lecture performance.
Here’s the top 10 individual lecture performances (out of 35 person/lecture entries) :
Brian Schwab – Rant 4.65
Dave Mark – Psychology 4.61
Rez Graham – Rant 4.60
Rez Graham – Panel 4.59
Matthew Jack – Tactics 4.59
Stéphane Bura – Rant 4.58
Ben Sunshine-Hill – Non-Traditional AI 4.57
Mike Robbins – Non-Traditional AI 4.57
Kevin Dill – Rant 4.57
Alex Champandard – Tactics 4.57
Highest Summit-Wide Performances
When you combine all the sessions people had, things shake out slightly different. These are the top 6 speakers for the year (out of 25). The number of lectures is in ( ):
Rez Graham 4.60 (2)
Matthew Jack 4.59 (1)
Alex Champandard 4.57 (1)
Ben Sunshine-Hill 4.57 (1)
Mike Robbins 4.57 (1)
Dave Mark 4.56 (3)
As with the lectures, there is a huge pack of people all clustered under this pack, but the people above are the ones that had broken away.
AI Summit Advisers Steve Rabin and Dave Mark want to thank everyone who was involved this year. While we have highlighted the people and the sessions above, we want to make sure that you all know that the entire Summit was incredible… again. Everyone who participated came well-prepared and the content was varied and insightful.
Remember that you can find the slides and videos posted on the GDC Vault. For your convenience, we will be linking to the individual presentations on our papers and presentations page here on GameAI.com.