042: It's time to try harder

[Before - Session 14] AI has raised the baseline for generic, average ideas

Artificial Intelligence is probably not going to solve for raking wet leaves out of garden beds each spring, and honestly—would we want it to? There’s a distinctly human appeal to gardening.

Now if AI could just solve for the ruinous impact of rabbits. (I am very close to acquiring some barn cats. If you have experience, please reach out!)

My MCAD students will be presenting ideas tomorrow for Assignment 2: Enhancing/retaining loyalty for a storied salon retail brand. I’ll have a recap Monday evening.

In the meantime, I’ve been researching and using a broad mix of AI tools to understand their impact on strategy and idea-creation, and refining a “Brand Narrative for CEOs in the Age of AI.”

One notion sticks out:

The bar for average creativity is now much, much higher

“Clutter” is getting much better, faster and cheaper.

There has always been a median majority of marketing content which does not set trends, does not intend to stand apart, deliberately looks like itself, and embodies the status quo. Cynics call it “clutter.” For the rest of us, this is very often our day in and out. There is nothing wrong with “average” content so long as it has integrity, craft, and a modicum of useful insight.

Every brand, agency and marketer makes some amount of clutter. Even the greats. This is the cost required to establish a creative baseline. If every single Insta ad, billboard, and packaging design was world-changing, that would simply become the new bar. Now try again, and try harder, please.

It’s incredibly hard to stand apart, much less sell it. So we churn what we can.

But now AI-generative tools have rewritten the rules for the baseline of creativity.

---

1️⃣ I built a marketing website in 30 seconds via Durable. Was it amazing? Of course not. But neither is most AI-generated material. The point was the speed. There’s zero excuse now if you need to move fast. For example…

---

2️⃣ The global architecture firm Zaha Hadid Architects just shared examples of how they leverage Dall-e2, Midjourney and Stable Diffusion to ideate and explore ideas—of which 10-15% move ahead. Studio Principal Patrik Schumacher said, “For me [AI prompting] is … very similar to verbal-prompting teams, referencing prior projects and ideas and gesticulating with my hands.” Note they use their own corporate name in prompting. This how you get to a greater quality and range of ideas to evaluate, faster.

---

3️⃣ All of these models were generated by AI, courtesy of LaLaLand. The baseline for prototyping across fashion, architecture, product design, packaging, advertising design and layouts is more enabled, and faster. You have zero excuse for not authentically representing your customer when you control the attributes for generating artwork.

---

4️⃣ Adobe Podcast (Beta): 🤯

I’ve been producing a podcast periodically over the past seven years with my father. Philip knows quite a bit about global choral music, running an opera company and being a church organist and choir director.

When we started, I used big microphones and complex editing software. This weekend I recorded our latest episode with an iPhone 14 (episode 22 goes live Tuesday), uploaded the file to Adobe’s speech enhancement process, and am finishing the edit in Descript.

The secret sauce here is Adobe’s software. They’ve figured out how to take “mildly okay” smartphone audio, clean up background noise and make the end result sound… really good? Here’s what an audio pro thinks.

And I can’t say enough for editing words, not audio waveforms. This UX shift is radically transforming how we create audio and video.

Key takeaway: It is now remarkably easy to produce very good sounding audio with little more than a smartphone and zero technical skill. Of course, you still have to have a brilliant idea, good taste, talented talent, and the wherewithal to bring it all to fruition.

---

5️⃣ Audio Pen: Solving an idea-making hurdle

Sometimes ideas arrive neatly. Most of the time mine are a haphazard stew of thoughts, scribbled onto anything at hand or (since the iPhone’s audio dictation got better), dictated into Evernote.

This second approach just got significantly more effective, thanks to ChatGPT and Louis Pereira. Here’s the explainer video.

Now, instead of dictating into Evernote, I dictate into Audio Pen. (3 minutes for free, 15 for paid/prime.) In the background, Pereira leverages ChatGPT to summarize my rambling into two concise paragraphs and a headline. You can adjust the settings from a one sentence summary to something much longer. You also control style, i.e. have it summarize “in the style of [favorite author].” Don’t like the initial result? There’s a re-do button.

I have no doubt we’ll see similar solutions inside Google/Microsoft/Apple OS soon.

Key takeaway: Idea initiation is often messy. GPT-infused dictation and summary makes it much easier to start.

---

6️⃣ Charisma as a Service (CaaS): Black Mirror en route

What if AI could remember faces, and prompt you accordingly? It’ll be commonplace, soon enough. According to Vice, a student combined GPT-4, speech recognition, and augmented reality to enhance face-to-face interaction. Forgetting faces will be a thing of the past.

---

7️⃣ Last but not least, productivity expert Cal Newport has done the world a service by explaining what’s actually occurring “under the hood” of ChatGPT. If you need a primer, this is a great place to start.