AI
Curiosity
Catalog

In 1994, most ad agencies thought the web was a fad. By 2000, every one of them had a "digital department." The ones who got there early didn't just survive — they led. The ones who waited mostly didn't make it in their original form.

The pattern is identical to what's happening now. A new capability appears. Most people dismiss it. A few lean in early. A few years later, the early movers are running the show and everyone else is catching up — or gone.

If you were around in the mid-90s, you remember when "building a website" felt exotic. Now it's table stakes. AI is at the same inflection point — and the window to be early is still open, but not for long.

Skepticism is healthy. There's a lot of noise around AI — overblown promises, bad demos, tools that don't work as advertised. You're right to question it.

The best way through skepticism isn't to be convinced by an argument. It's to use it for something small and real. Not a grand experiment — just: paste an email you're dreading and see what comes back. That's it.

Most people who describe themselves as skeptics about AI have never actually tried it on a real task. The ones who have — even one time, on something real — almost always come back. Not because it's magic. Because it's useful.

HAL 9000. JARVIS. Samantha from Her. The replicants in Blade Runner. Science fiction has been processing our hopes and fears about AI for generations — and now we're living in the part of the story where the technology actually arrives.

The interesting thing: most sci-fi got the emotional part right and the technical part wrong. AI isn't a robot overlord. It's closer to a brilliant, tireless collaborator who knows a little about everything and never gets defensive when you push back.

If you love stories — if you're drawn to ideas, futures, and "what if" — AI is genuinely one of the most interesting things to explore right now. Not because it's powerful. Because it's strange, and new, and nobody fully knows where it goes yet.

The real pattern isn't "AI takes jobs." It's "AI absorbs tasks." The jobs that disappear are the ones that are mostly made of repetitive, low-judgment tasks. The jobs that grow are the ones where human taste, relationships, and creative judgment are the core — not the wrapper.

At RedPeg, that means the executional grunt work gets faster and cheaper. The strategic and creative thinking becomes more valuable, not less — because there's more capacity to pursue it. Your job isn't disappearing. The low-value parts of it are, which frees up more time for the high-value parts.

The risk isn't that AI replaces you. It's that someone who uses AI well outpaces you. That's a solvable problem — and it's exactly why you're reading this.

When electricity became widely available in the late 1800s, most factory owners used it to power the same machines they'd been running with steam. It took 30 years for them to realize electricity didn't just replace the power source — it allowed them to redesign the entire factory floor.

The factories that thrived weren't the ones that swapped steam for electric motors and called it done. They were the ones that asked: "Now that we have this, what becomes possible that wasn't possible before?"

That's the question for AI. Not "what tasks can it automate?" but "what becomes possible now that wasn't before?" For a creative agency, the answer is enormous — faster ideation, better research, more time for the work that actually matters.

The most paralyzing moment in creative work isn't execution — it's the blank page. The moment before anything exists. AI doesn't replace the creative leap. It obliterates the blank page.

Ask Claude for 10 directions on a brief. They won't all be good. Some will be obvious. But two or three will spark something — a direction you wouldn't have found alone, or a combination of ideas that opens a whole new territory. That's the value. Not the output. The spark.

The best creatives aren't the ones with the most raw talent. They're the ones who generate the most ideas — because they know good ideas are a numbers game. AI turns that dial up by an order of magnitude.

Theory is fine. Results are better. Here are three things you can do in the next hour that will save you real time — no setup, no training, no special access required.

1. The email you've been putting off. Paste the situation into Claude. Get a draft back in 10 seconds. Edit it. Send it. Done.

2. The recap nobody wants to write. Paste your bullet notes from a meeting or event. Ask Claude to turn them into a client-facing recap. It takes 30 seconds.

3. The thing you need to understand fast. New client category? Competitor you've never heard of? Ask Claude to brief you in plain language. Faster than Google, and it synthesizes instead of just listing links.

In 1995, "social media manager" wasn't a job. Neither was "UX researcher," "growth hacker," "content strategist," or "data scientist" in any meaningful volume. The internet didn't just change existing jobs — it invented entirely new categories of work that became core to how agencies operate.

AI is already generating early versions of new roles: AI prompt engineers, AI creative directors who specialize in human-AI collaboration, experiential producers who design AI-integrated live events. In five years, there will be job titles that don't exist yet that will be standard on agency org charts.

The people who define those roles will be the ones who are curious about AI now — who are experimenting, learning, and building fluency before it becomes required.

The best analogy for working with Claude isn't "using a search engine." It's closer to having a brilliant intern who has genuinely read everything — every brief, every strategy doc, every marketing textbook, every piece of journalism published in the last 30 years — and can apply it to your specific situation in real time.

Like a great intern, you still need to direct it. You still need to apply your judgment to what it gives you. You still need to know what good looks like. But you never have to start from scratch, never have to do the low-value grunt work alone, and never have to say "I wish I had more time to think about this."

And unlike an intern, it's available at 11pm, never gets offended by feedback, and gets better the more specifically you direct it.

Claude scored in the 90th percentile on the Bar Exam. Top marks on the US Medical Licensing Exam. 95th percentile on the GRE. It outperforms most humans on standardized tests across law, medicine, math, coding, and science. This is not a search engine. This is a colleague with credentials most of us will never have.

And it's sitting in a browser tab you open twice a week to ask one question.

McKinsey research found knowledge workers using AI save 1.5 to 2 hours every day — not because the tool is magical, but because they actually ask it things. Consistently. For everything. That's a full extra workday every week, compounding.

At RedPeg, we have one of the most capable thinking partners ever built available to every person on the team. The question is whether we're using it like a colleague or like a calculator we pull out for specific problems.

One of the most common reasons people delay using Claude is the belief that there's a setup phase — that you need to "train" it before it's useful, the way you'd onboard a new employee. There isn't. There's no training. There's just asking.

Claude already knows experiential marketing, RFP writing, client communications, brand strategy, event production, and creative development. Not because you taught it — because it was trained on more relevant material than any person could read in a lifetime. You bring the context of your specific situation. It brings the capability.

The "training" people think they need is actually just good prompting — and good prompting is just being specific. Tell it what you're working on. Tell it who it's for. Tell it what format you want. That's the whole setup. You can learn it in one conversation.

Imagine someone handed you a laptop in 1985 and you said "I'll learn to use this in my spare time" — then kept doing your work by hand while the computer sat in the corner. That's what treating AI as a side skill looks like right now.

AI is not a software tool you add to your stack, like learning a new Asana feature. It's the medium your work should live in. 90% of what you do every day — writing, thinking, researching, planning, communicating — Claude can do faster alongside you. Not instead of you. Alongside you.

The people at other agencies integrating this now are not smarter or more technical. They just started earlier. The gap compounds. A person saving 90 minutes a day has an extra 7.5 hours a week — nearly a full extra workday — compared to someone who doesn't.

No more reading about AI. No more demos. Open Claude in a new tab. Pick one of these — the one that costs you the most time today — and just go.

The RFP you're dreading. Paste the brief. Ask Claude to draft the executive summary. 45 seconds.

The recap you've been putting off. Paste your bullet notes. Ask for a client-ready version in your brand tone. 30 seconds.

The client you know nothing about. Type their name and category. Ask Claude to brief you like you have 10 minutes before a call. Better than Google.

The email thread that's been sitting there. Paste the whole thread. Ask what the three most important action items are and draft the reply. Under a minute.

The moment you get a result that surprises you — that's the switch. Everything changes after that moment. You are one task away from never working the old way again.

Some people have seen the demos. They believe Claude is capable. They've even used it and gotten great results. But it's still not the first place they go. When a task lands on their desk, the instinct is still to open a blank doc and start — not to open Claude first.

That gap isn't about knowledge. It's about habit formation. The tool isn't wired into the reflex yet. And habits don't form from information — they form from repetition plus immediate reward.

The fix is simple but requires a commitment: for the next two weeks, Claude is your first move on every task. Not sometimes. Every time. Even the small ones. Especially the small ones. The reflex builds through repetition, not resolution.

Within two weeks, opening Claude will feel like opening email. You won't remember deciding to use it. You'll just do it — because the output is better and you've proved that to yourself enough times that your brain stopped arguing.