How Artificial Intelligence Improved Our Virtual Conference

Written by Arianna Rehak on July 2, 2018

What comes to mind when you think about Artificial Intelligence? Is it images of robots, machinery, and science fiction movies?

In fact, AI has already made its way into our homes and offices, and that is great news, because it is successfully automating processes and taking care of the most arduous of tasks.

With that in mind, I wanted to share some ways that has leveraged technology to help us develop content efficiently for our semiannual virtual conference, SURGE.

Here are some of the most promising AI-powered innovations that we’ve been working with:

1. To more efficiently write articles.

A simple and ever-improving tool we use constantly is voice-to-text technology. If you’re looking for an example, most phones have this functionality built right in, and they’re getting really good! If you check it out, you’ll notice that the text will often correct itself. Even if it thinks you said one thing, with further context, it actually adjusts itself! Also, try telling your phone to add a comma in the sentence, or a “smiley” at the end, and it will!

We often use voice-to-text to take video content that we’ve captured for the conference, and share it through a different medium. For example, we’ve found that an hour-long interview can provide excellent fodder for written content. The first step of transcribing the text is a lot faster through voice-to-text, and there are programs out there that will actually do this for you automatically (although they may require some manual editing!)

2. To capture video discussions asynchronously

An entire session of our first conference was actually put together with the help of AI. Due to an unusual number of family emergencies and conflicting timelines, the speakers of one particular session were not all available to record their session at the same time.

The options were to cancel the session, or to get creative.

So here is what we did: the first speaker recorded himself giving his thoughts on the topic as a base. This was then transcribed through voice-to-text, and the second speaker was given the transcription. Together we identified various areas where it made sense to jump in, and recorded his response. We then brought the revised transcript from the first two speakers to the third and fourth speakers.

This is an extremely time-consuming process, but the voice-to-text technology significantly streamlined it, and you might find yourself in situations where it comes in handy!

3. To analyze what our audience values

Our third use of AI was entirely different: we had fun playing around with IBM Watson’s free tool. Insert a text into the tool, and Watson will analyze the conversation. We found the keyword analysis function the most useful. We took the text from the attendee chat of SURGE, and dropped it into the keyword analysis. Essentially, Watson pulled out the key expressions that were used based on frequency and significance of use. If you have a large enough data set of text, this is a really interesting way of understanding what your community values. Our keyword analysis was useful in determining future content on our digital publication, and even future SURGE sessions.

Be careful though! Context is important. For instance, in one of the sessions, Star Trek was used as an example. This prompted a tangential discussion in the attendee chat about the show. If we had taken this at face value, we might be holding a completely Star Trek-themed conference session!

There are plenty of free and inexpensive tools available to help improve efficiency, learn about your audience, and serve them better. As we experiment with new forms, we’ll be sure to share!