121: Let's get weird
[During - AI for Artists and Entrepreneurs] Leadership + creative AI updates
Last night was the antepenultimate session of the AI for Artists and Entrepreneurs class at MCAD. We worked on our final projects while listening to Rick Webb’s Abbreviated (Taylor’s) Poets Dept. playlist. And we cycled through most of the AI+Creativity Update content below.
The students are working on a diverse range of ideas for their projects—each of which must utilize at least two different GenAI tools. They’re making TV commercials, product packaging, a video game, an Etsy store, a gaming “lore” keeper GPT, fine art, and more. I’ll be sharing their work in mid to late May once class wraps.
In the meantime, let’s talk about the Boss.
Leadership gets weirder in the age of AI
It wasn’t that long ago you could ignore software and still lead a company successfully—never mind be wildly successful being wildly creative. (Anyone else listening to the Acquired podcast episode about Microsoft?) Even when you and your firm ultimately succumbed to software, it took years to implement, and most important—it took years for culture to change. The Internet, smartphones and social sped things up… eventually.
But leadership in the age of AI is and will become increasingly weirder.
Last September various academics including Wharton’s Ethan Mollick released a research paper developed with Boston Consulting Group. They measured task output quality across 750+ consultants with a range of experience from brand new to 30 years on the job. Those with access to AI performed better. On everything. At every level of experience.
Consultants using AI finished 12.2% more tasks on average, completed tasks 25.1% more quickly, and produced 40% higher quality results than those without.
What is it about this new technology, this moment, that separates AI’s impact on leadership from previous seminal moments (software, Internet, social)?
So much of leadership can be summed up as a keen sense of the patterns which worked best before, and those patterns which didn’t. And leaders typically had an advantage of access to patterns. But as Rishad Tobaccowala writes, “Talent can find information, opportunity, and knowledge at a click of a mouse, a sharply written prompt or the pinch of a finger.” While the Internet and social increased access, what AI does is enable and spread that “keen sense.” AI doesn’t just provide information, it helps an individual discern; it is designed and operates as an adjunct to your thinking.
How do you lead teams which—because of AI—might not need traditional forms and expressions of leadership?
Let’s try philosophy.
Marco Argenti, Chief Information Officer at Goldman Sachs, penned an op-ed for the Harvard Business Review (thanks for the tip, Greg!) last week called “Why Engineers Should Study Philosophy.”
The quality of the output of a large language model (LLM) is very sensitive to the quality of the prompt. Ambiguous or not well-formed questions will make the AI try to guess the question you are really asking, which in turn increases the probability of getting an imprecise or even totally made-up answer (a phenomenon that’s often referred to as “hallucination”). Because of that, one would have to first and foremost master reasoning, logic, and first-principles thinking to get the most out of AI — all foundational skills developed through philosophical training.
I got to talk about AI with a room of CEOs two weeks ago. The feeling “this time is different” was palpable. They are seeing a new kind of enabled employee take shape. There’s a weirder definition of leadership emerging—one which allows, and encourages, far greater ability and agency downstream from the c-suite than previously supported.
And it’s occurring much faster than we’re used to.
AI+Creativity Update
🤖🇱🇧 Lebanon’s parliament hasn’t been able to elect a president for two years due to political divide. So, AnNahar, Lebanon’s oldest newspaper, built an AI President based on 90 years of impartial journalism. Then the paper’s editor-in-chief, Nayla Tueni, interviewed the AI President on live TV. Kudos to the newspaper, agency IMPACT BBDO Dubai and AI production firm Addition for giving it a try.
🤖🕥 The Dali Museum launched the Dali AI Lobster Phone. Now you can “talk” with Salvador Dali about his work if you’re at the museum. You’re going to see more and more of this kind of project—an organization leveraging its unique data to present a novel experience expanding upon a known context.
🤖🎤 The 2nd annual BRXND AI Conference is May 8 in NYC. Last year’s event and its Ad Turing Test generated lots of keen insights. You should go!
🤖🥣 Campbell’s Canada and Zulu Alpha Kilo re-interpreted the infamous soup can in the style of other art movements. It’s an ad stunt. Wish they’d tackled a 72-page prompt playbook like Dove. Or maybe incorporated artist perspectives from the Hello History AI app.
🤖🤔 Meta launched its latest AI model, Llama 3, inside Facebook Messenger, WhatsApp and Instagram. You can also access it via meta.ai. Casey Newton wrote an extensive review. Now we get to see what happens when billions of social users have free access to GenAI.
I’ve only spent a few minutes testing Meta’s tool—it generated a SWOT analysis for me, but refused to cite sources, and then it claimed it had not generated a SWOT analysis (but had actually done so, twice 🤦🏻). That said, the image generation tools are interesting. And fast! Check out this video of a real time test, and notice how the tool pops up a reference image almost as soon as I’ve finished typing the word.
🤖📱 And Google just launched new GenAI capabilities for advertisers inside its Demand Gen tooling for YouTube and Gmail. “For example, if you have an outdoor lifestyle brand that sells camping gear, use prompts like ‘vibrantly colored tents illuminated under the Aurora Borealis’ to create images that will appeal to those who are shopping for camping trips to Iceland.” You knew this was inevitable.
🤖⛏️ I’ve been digging through the wealth of presentations from Google Cloud ‘24. This slide from “Become an expert in advanced, multi-turn prompting in Gemini for Google Workspace” resonates if you’re struggling to understand how to structure prompts:
🤖📸 Microsoft Research (Asia) introduced VASA, a generative AI approach which needs just one image and an audio clip to produce video with, “lip movements that are exquisitely synchronized with the audio, but also capturing a large spectrum of facial nuances and natural head motions that contribute to the perception of authenticity and liveliness.” Amazing. Also concerning! They clarify Microsoft is, “opposed to any behavior to create misleading or harmful contents of real persons.” As AI Tool Report explains, Microsoft, “has no plans to release an online demo, API, or product, preferring to make sure the model ‘will be used responsibly and in accordance with proper regulations.’” We’ll see, won’t we?
Example: