117: AI is about making decisions
[During - AI for Artists and Entrepreneurs] Jon Stewart and Peter DuCharme weigh in
I read books the old fashioned way, that is to say with pen in hand, and slowly. I’m at page 92 in Ethan Mollick’s Co-Intelligence, and I’ll be hosting a book club discussion at the end of this month. A few of you have expressed curiousity. Maybe there are more who will join?
Meanwhile, Mollick was interviewed here for BSG, and in this video for Technovation.
But wait. Jon Stewart has opinions about AI!
I’m pretty sure Mollick didn’t write his book for laughs. Jon Stewart definitely wrote and performed this segment for laughs. (Jump to 06:08 for the meat of it.)
If you’ve only consumed headlines, and find technology generally unsettling, then you are Stewart’s audience here. He plays on all the AI tropes, while demonstrating little comprehension of what generative AI actually is, or does. “Types Questions Guy” serving as proof. Granted, maybe that’s his actual intent—to demonstrate how naive, silly, and untrained AI leaders and government leaders appear when reduced to soundbites. They do all sound quite ridiculous, Stewart included. His own naivety is similar to Sam Altman’s take on the function of marketing.
Reductionist, non-curious rants stand in stark contrast to what people like Mollick are attempting to convey. My biggest takeaway is Mollick’s definition of AI; that it is:
“A Prosthesis for Thinking”
☕️ Just like coffee?
In my experience, generative AI is much less about a binary response, like a search result or calculator answer or Jarvis making toast, than it is a means of eliciting better ideas. It is a companion for solving problems. As example, what will the next episode of the Renaissance Man podcast focus on? I want to illuminate the ways in which the role and output of a founder/music director for a 50+ choral arts non-profit is akin to an MBA-trained corporate executive.
So ChatGPT and I had a productive dialogue—I outlined my objectives and it suggested a course of action. Some of it I agreed with, some disagreed; I rerouted the conversation twice, and while the technology did not write my outline, it did help me figure out a shape for the interview and gave me some useful references to explore. GenAI helped me decide.
Thank you, Peter DuCharme
“Create like Pollack, but revise like Albers.”
In other words, “the most important thing is critical thinking and listening.” That’s the gist of GenAI according to last night’s guest in our AI for Artists and Entrepreneurs class, musician and composer Peter DuCharme (LinkedIn). It’s as much about knowing what to throw away, as what to curate and keep—and why. “As long as you’re making decisions, you’re an artist,” says Peter, especially within the realm of GenAI. What you select, how you bring your opinion into the work, is what really matters.
Music remains something we all enjoy, but very few of us create. Unlike writing, or even image-making. What’s encouraging in this era is the enabling of creation, building upon samplers, DAWs and the like—now if you can type, you can compose. And this alone enables a different kind of problem solving. As Peter continues, “the best kind of communication is a reference.” The potentially troublesome practice of, “make it sound like [famous song]” might be mitigated by tools which demystify the worlds of harmony, melody and rhythm.
Of course, the world of music is somewhat unique when it comes to AI if only because we’ve sort of been-there-done-that with a long history articulating and defending artist’s rights. As I mentioned before, “the entire [music] system is built around compensation for output. Generative AI is all about input. And there isn’t a financial model for that. Yet.”
So there’s a worthwhile exploration within GenAI music beyond the creation of audio. Peter works for Boomy, and their point-of-view incorporates distribution—imagine if Midjourney enabled your GenAI work to syndicate automatically into newspapers or magazines. I suspect we’re going to see new and fresh ways to think about music in the contexts of applications and content creation (see the note about Google Vids below).
I’m very grateful for Peter’s insights and encouragement with the class. It’s a thrill to reconnect with an old band mate under a new but related context. Peter still writes and produces great alt-rock songs. You should listen.
AI+Creativity Update
🙋🏽♀️ Meet Souki Mansoor, a Creator Community Specialist for OpenAI. She gave a presentation on storytelling and filmmaking in the age of AI for the Future Proof Creative Community.
🤖 You can now edit images generated in OpenAI’s DALLE3 (via ChatGPT Pro or free via MSFT Copilot). Very similar to the “vary region” functionality in both Midjourney and Firefly. Sometimes it works.
📹 “Google seems to want to signal that this is something more than just another way to present a deck.” Better late than never? And maybe creating video is one step more complicated, still, than creating a deck? Anyway, The Verge reports on Vids, coming this summer. Sounds like Loom, and Vimeo, but within Google’s system minus YouTube. And with AI. Of course.
😉🤖 “Human smarts and silicon smarts work in very different ways.” This Clive Thompson piece from 2010 resonates with a lot of what we’re learning now about how to get the most from generative AI. (Thanks, Robert!)