105: I labor, gladly, in the curiosity and persistence business
[During] We covered Politics, Text-to-Video and Social Media in classes this week
Yesterday Seth Godin noted the first hundred of his blog posts “were read by fewer than a dozen people.” 🤔 Now imagine deciding every day, as Seth puts it, “not to stop.” Maybe the gap isn’t nearly as wide as we think it is. Or as Anthony De Mello says, “My business is to do my thing, to dance my dance.” Onward!
[Agenda]
On Monday the AI for Artists and Entrepreneurs course dove into generative video. On Tuesday the Persuasion & Marketing course focused on politics, with deeply insightful guest Zach Rodvold. Then on Wednesday I subbed for my friend Ben to lead MCAD’s Agency class to visit The Social Lights. Did I mention I’m also building a curriculum on short form writing? Then there’s the freelancing work. The point is to not stop, to keep dancing, I guess.
They gave us socks. And cookies.
[Wednesday]
The Agency class is simply that: Visit a different agency every week. This week was TSL. Last week was KNOCK. The week before, Carmichael Lynch. The point is to compare and contrast, to realize the differences between advertising-driven, design-focused, tech-designed, or social-oriented approaches to being an agency; and to encourage students to network.
I’m so grateful we have such a diverse and generous range of agencies here in the Minneapolis / St Paul metro. I mean, the team at TSL didn’t just show an agency real and wrangle a few talking heads. First they handed out socks. Then they brought out young talent, people close to our student’s age. And they revealed the truth bridging a college experience into an entry level position, and the realities of the day-to-day work. They made the experience appealing, intriguing and worthwhile.
Then they gave us cookies.
From Common Sense to the Dean Scream Meme to Misinformation
[Tuesday]
The most surprising thing about persuasion and politics is how little has changed since Thomas Paine’s Common Sense in 1775. A huge thanks to Zach Rodvold for illuminating several hundred years worth of strategy, framing, signaling, technology, culture and luck that encompasses the use of persuasion within politics. Especially technology. From Lincoln’s use of newspaper op-eds, to FDR’s embrace of radio, Kennedy recognizing the power of TV, to Obama and digital, and now AI there’s a compelling history of media tech influencing the political realm. And luck. As Zach tells it (from his position within the campaign), if the audio had just been room sound and not a direct mic to TV, none of us would have heard Howard Dean’s infamous scream, or thought much of it. Dean was yelling because the room audio wasn’t loud enough.
We also watched journalist Steve Richard’s take on politics, which is that it’s the issue of humans. Richard’s point is, “the public's inability to see politicians as human beings is the perfect breeding ground for comedy and tragedy to collide.” And we benefited from John Green’s Crash Course in the history of U.S. politics, in short—someone made a musical about most of it.
So the challenge for students (and frankly all of us) today is in recognizing the system at work, being aware and wary of the tools of persuasion—they’ve always been there, but have grown increasingly more subtle, harder to discern, and washed amidst a deluge of information.
What will you do with this infant technology?
[Monday]
The AI for Artists and Entrepreneurs class is close to wrapping up our introductions to text, image, video and audio technologies. First we learn the tools, then we put them together in different contexts.
The key challenge with text-to-video, I think, is actually having a valid use case or assignment. In the business or presentation world, we tend to speak with words (thank you, ChatGPT) or static images (much appreciated, Midjourney). Sure we might bake in a meme video, and yes we enjoy the endless stream of distraction on our devices—but far fewer of us create video for work. So we lack a context, a reason (and experience), to leverage text-to-video.
Is video additive to the message or experience you’re trying to convey? I’d argue most marketers, most agency workers—the very people generative AI should appeal to—have little experience answering that question.
Oh, and today’s text-to-video tools are outputting quality that’s loosely comparable to last year’s text-to-image results. In other words, text-to-video quality isn’t quite ready for prime time. (But maybe Sora changes that?)
So what should I generate?
After we explored the UX and processes inside Runway, Pika Labs, D-ID and HeyGen, I gave the class two different prompts.
Create a symbolic video expression of freedom
Create a talking head explaining how to use the hydrocoptic marzlevanes of the turbo-encabulator
The first asks us questions—what does “freedom” mean and what might an expression of it look like? This is more in the realm of fine art than marketing. But it helps us think about prompting and output differently than the second assignment. This assignment demonstrates the value of tools like Runway and Pika Labs.
The second is more pragmatic, and speaks to the strengths of tools like HeyGen and D-DI: We need an FAQ video (based on an old engineering joke).
Here’s a supercut of the student’s output for an “expression of freedom.”
And here’s how a handful of them leveraged text-to-video tools to create a talking head FAQ.
Next week we’re going to explore AI generative audio tools. And after Spring Break, we’re going to start putting everything together in various contexts, from fine art to branding to performance marketing.