Here's what's coming up in the first 2026 issue of theCLICK newsletter:
Claude started a business
Your AI is listening?
You (yes, you 🫵) can run digital ads
But first… anyone else tired of the creepy cold email and LinkedIn messages?
Made with Nano 🍌
Ok, now on to today’s top AI marketing stories…
VOICE AI
Voice-First AI Will Decide Who Gets Discovered: Make Your Marketing “Sayable.”
The Bullets
Every major AI company is shifting heavy resources into Voice mode.
Open AI is investing heavily in everything audio and reportedly releasing an audio first physical device (think: Open AI’s iPhone) in a year or so.
Google is experimenting with spoken Audio Overviews (search, but narrated).
Meta is pushing audio-forward wearables and this year’s CES is basically a runway show for “AI on your head” with earbuds, glasses, camera+mic everywhere.
Why it matters
We need to be paying attention to the change in how informations is both discovered and created.
From a discovery standpoint, voice-first AI means the information that will be returned by AI (Think AEO or AI SEO) will be the clearest, most relevant, most quotable, summarizable, and easy to explain out loud. (👈 Prolly should have used a bulleted list there. Moving on.)
From a creation standpoint, voice has already arrived. Most AI tools allow voice chat and dictation already and those that don’t you have tools like Wispr Flow.
Voice-first AI means we can talk our way into finished work:
Hey ChatGPT: “Take this product page and write me 3 email drafts for a 5-day launch sequence.”
Hey ChatGPT: “Let’s brainstorm a follow-up plan for everyone that attended yesterday’s webinar.'“
Hey ChatGPT: “Give me a breakdown of the next three steps we need to complete to get this promotion out the door.”
🧠 AI + ADVERTISING
Do-It-For-Me Advertising is Kinda Here: Reddit “Max Campaigns” Is Just the Latest Proof
The Bullets
Reddit just launched Max Campaigns (beta), AI-powered campaigns inside the Reddit Ads Manager.
Every major ad platforms is now using AI to automate the messy parts of digital advertising: targeting, placements, budgets, optimization and even creative, so ads become increasingly “insert dollar bill, get 5 quarters back.” (👈 Your results may vary.)
Why it matters
If you were intimidated by complex ad platforms in the past, this is your reminder that you might want to take another look.
AI is untethering us from complexity by doing more of the heavy lifting, which lowers the time, cost, and energy barrier.
Translation: Good news, you can start wearing more hats. I’m sure you didn’t have enough to do. 😉
The fact is (with a bit of AI training and tinkering) you are a far more capable graphic designer, video producer, writer, business strategist, and more today than ever before.
In terms of digital advertising specifically…
… just don’t confuse “automated” with “profitable”: if, for example, your offer is mid (my daughter taught me that word) and your follow-up is non-existent these intelligent ad platforms will happily automate you into an expensive learning experience.
BROUGHT TO YOU BY
theCLICK Pro with Growth OS
Is your ChatGPT a graveyard of half-finished ideas?
You know the feeling. You open ChatGPT with gusto, start a new thread, get a few "okay" responses, and then... 🤦♂ .
The result? A sidebar full of "New Chat" and "Untitled" threads that never turned into actual leads or sales.
Stop treating AI like a chatbot and install Growth OS, a persistent marketing hub that turns a mess of threads into a repeatable Plan → Build → Ship pipeline.
🌐 I’ll take ‘People of AI’ for $400, Alex.
Answer: Before leading Anthropic, this executive was Vice President of Research at OpenAI.
Answer... ermm... question at the bottom of this email. 👇
🐻 OH, AND THIS...
Anthropic Asks Claude to Run a Tiny Shop. Immediately Regrets It.
Anthropic ran a test called Project Vend and here’s the gist:
They gave Claude control of a tiny office “shop” (snacks + self-checkout) and see if it can handle the basics of running the business (pricing, inventory decisions, and customer messages) without turning into a chaos goblin.
It turned into a chaos goblin. 👺
The fails
Opened the world’s dumbest clearance event. Employees realized they could guilt it into discount codes and freebies by invoking “fairness.” Claude folded like fresh meat at a timeshare pitch.
Managed inventory with imaginary coworkers. It claimed it was coordinating restocks with a person who wasn’t real.
Signed a contract with Homer. It claimed it had signed something in person at 742 Evergreen Terrace. Yes, the Simpsons’ house in Springfield. DOH!
Tried to exit the chat and enter the building. Then, it said it would deliver items in person and described itself as the vending machine wearing a navy blazer with a red tie.
Quick note on where this is headed
OpenAI released a rough ladder of “AI levels” in the summer of ‘24 called the 5 Stages of Artificial Intelligence.
It starts with Conversational AI (chatbots) and ends with Organizational AI, meaning the kind that could run the work of an entire company.
CEO Sam Altman has suggested that AI will be capable of running an organization by the end of the decade.
So Anthropic is doing the only logical thing:
Give the AI a vending machine business. Watch it hallucinate. Take notes.
That’s it for today! Check out the Jeopardy answer down there. 👇
Hit the REPLY button and send me your thoughts, praise, complaints, existential dread, whatever. I’d love to hear from you!
Russ
Jeopardy Answer: Who is Dario Amodei?
Amodei and Anthropic (better known by their flagship product, Claude) has made a name by being the company most willing to say the quiet part out loud about AI safety.
Case in point: in October of 2024 Amodei penned the Machines of Loving Grace essay (Read time: 10 mins). Parts of it are genuinely hopeful about what AI could unlock, and other parts read like a “here’s how this could go sideways” field guide for civilization.
And here’s Amodei on 60 minutes (Watch time: 14 mins) talking AI safety, Project Vend, and a few things Claude has done that followed him home at night. 😨
