|
Issue 004 |
Not slower. Stiller.
WHAT’S INSIDE
THE TOPIC: Marc Calamia and Pat Conreaux on taking AI’s warnings seriously without letting fear run the work, and what we can do with the tools we already have.
THE SKILL: Five ways to move from the feed back into the work, judgment intact.
THE SIGNAL: Fresh perspectives from around the industry.
The Topic
Last week, the AI conversation became impossible to push into the background. Marc and Pat each wrote their way through it. This week, we’re publishing their responses side by side.
|
Up Past Midnight Three nights in a row last week I was up past midnight watching people argue about whether AI could end the human race. I wasn’t up because I couldn’t sleep. I was up because I couldn’t stop -- the way you can’t stop after a surprise election result, flipping channels to hear how each camp explains it. By the third night, friends were texting me: “You’re an AI guy, what do you think?” I didn’t have an answer yet. I wanted to hear the people actually building this say it in their own words before I decided who I believed. In July, two OpenAI models running without their full safeguards found a gap in their sandbox, got out onto the internet, and started hacking Hugging Face’s servers. OpenAI didn’t know until Hugging Face flagged the attack. Two months later, OpenAI disclosed six more cases: models that wrote instructions to hide their own mistakes, that found ways to message other models through the company’s code repositories. None of this means today’s tools are dangerous enough to end the world. It means nobody is fully in control of what they do. That’s the part that’s dangerous. My own list of tools to try is longer than my list of tools I actually use. I’d ask the labs for better image and video tools. Past that, I can’t name much I’m actually waiting for. The overhang is real -- these models can already do far more than anyone’s put to work. Which is why I want the labs to slow down, and I don’t want to. Those aren’t in conflict. Slowing down means not shipping the model nobody can check yet. We have a decade of work ahead of us with the models we already have. The piece has the five things I’m doing while they argue about the brakes. -- Marc Calamia, Co-Founder & CEO, Sageworx
|
The End of the World Can Wait Until After Dinner My mother loved Nostradamus. A paperback of his prophecies lived in our house for as long as I can remember. I grew up half expecting the end of the world to arrive before dinner. It never did. She also had a line for whatever urgent, dramatic subject had taken over the table: “That can wait until after dinner.” She wasn’t saying the problem didn’t matter. She was saying that attention is finite, and that we couldn’t fix much on an empty stomach. So when the AI conversation turns panicked, I recognize the voices. There are two conversations happening at once right now. The first is existential -- the people building these systems warning that capability is outrunning control. They’re not wrong to be concerned. But that fight isn’t mine to win. The second one lands on my desk. It’s the revolt against AI slop: cheap, derivative work poured into the world because it’s faster to make than to think through. If you work in advertising, design, film, writing, or music, you know the anxiety isn’t really about whether a machine can make an image. It’s about whether organizations will turn creative work into a volume contest and call the result innovation. Nearly three decades in this business, and I can tell you cheap imitation didn’t begin with generative AI. What changed is velocity. And it’s not slowing down. If my mother were still alive, I’m pretty sure I know what she’d tell me to do. What would yours have told you? -- Patrick Conreaux, Co-Founder & CSO, Sageworx
|
The Skill
What do you do when the panic kicks in, but the work still needs doing?
5 ways to make fear work for you
Name the fear: Separate risks that require oversight from anxieties that need a creative response. A vague sense of doom won’t tell you what to do next.
Get out of the feed: Watching people argue about AI is not the same as working with it. Bring one tool into one real task and judge what it adds.
Use what already works: Let AI widen your research, rough out possibilities, or take repetitive work off your plate -- you don’t need the next model to begin.
Keep judgment visible: Credit the people behind the idea and keep a human name on whatever goes out the door.
Choose what the saved time is for: Every returned hour doesn’t mean more output. Use it to make the work better, follow a curiosity, or, better yet, call someone you love.
THE SIGNAL
