This transforms the road.
— Horton Freire Bot (@hortonfreirebot) December 14, 2016
This will be my last blog post on the series I’ve been doing over the book We Make the Road by Walking, which included three chapter posts (1, 2, 3), a technical post on my quote generator, a co-authored post with Amy Collier, and a Twitter bot. In the end, I’ll have written roughly 12,000 total words on #HortonFriere, so many thanks to Bryan Alexander for organizing this online book club.
HortonFreire was the right opportunity at the right time. It became, more than anything, a way for me to channel my energy towards something that felt productive within my professional community. Recently, Alan Levine wrote a blog post about building a neat animation web tool based off of my Hypothes.is annotation spreadsheet, which pulls in all of my book highlights. I replied with this comment:
I’m thinking more about building as a way of processing or as a way of contributing.
I like when I get to both contribute and stretch myself through little projects. It’s only icing on the cake when someone as talented as Alan builds on top of it. There should be a name for all projects that Alan and I both contribute it. I’m still mulling it over but I like the sound of “Adam Levine” at the moment.
Doing the co-authored post was also one of those moments and probably the most ambitious project I did within #HortonFreire. Please take the time to read it if you haven’t. In one fell swoop, we covered openness, scaffolding, grief, politics, participatory inquiry, and the future of infrastructure–aka our “roads.”
I was really connecting with Amy and her blog posts so I DMed her about having a virtual call in a similar format to the book and releasing it as a blog post. It gave me an opportunity to ask Amy some questions about her thoughts on the book while also thinking through written conversation as a medium. John Stewart and I were having a conversation about what’s the biggest difference between a podcast and a book like this. I have to admit that I’m a real big fan of the conversational style if only because it makes something like “critical pedagogy” much more approachable. It de-academicitizes something that I’ve come to think of as a very complex topic.
Amy and I used Skype for the phone call and I used a tool that David Kampmann turned me on to called Call Recorder for Skype, which makes a stereo recording of both inputs and outputs. I then took the audio file in to Audacity so that I could dial back the tempo which makes it a little easier to transcribe.
I will say that it’s pretty funny to listen to your own conversation at 70% speed. Eventually you get use to the slow pace of it and forget that you’ve sped it down. It mostly just sounds like you are just really struggling to find your words.
Our conversation was right around 45 minutes long at full speed and it took me maybe 90 minutes to transcribe it all in a Google Doc. I then gave us a few ground rules which was to only lightly polish the transcript although ach party was allowed to rewrite a single sentence if they wished. If you read it, you’ll notice that the sentences aren’t always perfect but that’s the point. I really was hoping it would reflect the natural flow of a conversation and I think it lives up to that.
Reflecting on this now, I do feel like we rushed through a lot of moments in order to fit everything into the 45 minute time slot. There are a couple of areas where we probably could have expanded on a bit. But I can now see how these books can come together rather quickly. I can imagine it would only take a handful of three hour conversations to build this.
The next question was where to publish it. We both agreed that a neutral space would be best. I found a tweet that said you can co-author Medium articles that are in Medium publications, but this seems to not be true as I never figured out how to do that. I was also curious in the new bold.io tool which is a publishing platform that doesn’t require a login, but that still is a bit too beta for me to trust. We landed in using Github pages and a Jekyll Boostrap theme.
What’s neat about Github pages is that anyone could fork the site, add their own blog post, and do a pull request to our repository. This means that anyone can add their own conversations as blog posts to the blog if they so chose to do so. It gives some kind of visual of what a federated social media network of transcribed conversations could possibly look like. Because there’s probably a market of like 11 people looking for just that.
I also made my first Twitter bot. Once Alan had released his project, I had the itch to build one more thing. It felt like a natural progression to go from quote spreadsheet, to quote generator, to Alan’s full blown webpage, to a textbot. I had a read a couple articles about Twitter bots that were Python-based. I’m still inimitated in that area so I went hunting for an easier solution. Thanks to Tom Woodward, I’ve become addicted to the Google Spreadsheet to Power Everything model and googled “Google Spreadsheet Twitter Bot.
As luck would have it, I landed on a post from Zach Whalen from Mary Wash. Zach has his students create Twitter bots with a Google Spreadsheet. There are four different flavors of bots you can create including one that pulls from your Twitter profile via Martin Hawksey’s TAGS tool as well as a Markov chain algorithm text generator, which has you paste in a text corpus and then puts together a sentence string based on pairs of words in the text. One of the neater pieces of the tool is that you can generate a preview that will show you potential tweets, so I pasted in my collection of now 100+ hypothes.is annotations of We Make the Road and hit go.
Terry Elliott, who was actually the person to invite me to first annotate the text using Hypothes.is, made a comment about whether you could tie the bot directly to the Hypothes.is tag.
— Terry Elliott (@telliowkuwp) December 12, 2016
This seemed like an intriguing idea, so I took Zach’s code that was tied to Hawkey’s TAGS tool and rerouted it to my Hypothes.is spreadsheet. I did indeed get it to work although for some reason the sentences are nearly as complex.
So, for now at least, I’m sticking with the Markov generator. Feel free to follow @HortonFreireBot for your daily #HortonFreire zen (thanks @HortonFreirebot for writing the title of this blog post and other things I agree with)
But I say revolutionary.
— Horton Freire Bot (@hortonfreirebot) December 15, 2016
I don’t know what it means when you decide that you believe bot gobbly gook. But I’m at a point where I don’t know what sources to trust, so create your own.
I’m excited to follow more posts as others continue their way through the book. While this is my latest “official” post, I’m excited to participate in an ongoing conversation. Now alls I got to do is write a couple end of year blog posts. What’s it mean when you are ready for 2016 to be over but you also don’t want 2017 to come?