📈 Reddit’s ad business reported a 69% YOY revenue increase in Q1.

✈️ A Spirit Airlines fan has launched Spirit 2.0, a crowdsourcing effort to buy the airline.

🎧 Spotify identifies human creators with “verified” checkmarks.

🤖 Data: Gartner finds 40% of agentic AI projects will fail.

📊 Report: Google Ads CTR is rising, but conversions remain flat.

🎓 Syracuse University launches Creator Economy minor.

New research from SparkToro highlights a huge AI misconception: While we tend to think of tools like ChatGPT as consistent “answer engines,” they actually behave more like probabilistic generators.

In controlled experiments with nearly 3,000 responses, identical prompts rarely produced the same list of brand or product recommendations (less than 1% of the time) and almost never in the same order.

We’re used to traditional SEO, which assumes stable rankings. But AI outputs appear to show variability in three dimensions: which brands appear, how many are listed, and how they’re ordered. So, “tracking your position” inside AI responses is pretty much meaningless. It’s not fixed.

But while rankings are chaotic, brands that consistently show up, regardless of position, are more strongly associated with a given intent in the model’s training and retrieval patterns.

Here’s what to know:

• AI recommendations are non-deterministic. The same query yields different outputs almost every time.
• Rank tracking is unreliable, as order and inclusion constantly change.
• Prompt diversity amplifies chaos. Users phrase the same intent in wildly different ways.
• Visibility share (frequency of mentions) is a more defensible metric than rank.
• AI tools reflect association, not endorsement. They predict likely mentions, not “best” choices.
• Marketers should demand transparent methodologies from AI tracking vendors.

Head to SparkToro for a closer look at the numbers.

NapoleonCat is a helpful social tool for managing conversations at scale and pulling comments, messages, and reviews from across channels into one dashboard.

Beyond the basics (scheduling, collaboration, analytics), there is AI-powered moderation to filter spam, hide harmful comments, and auto-respond to common questions.

It treats social media less like a broadcasting channel and more like a support channel.

Don’t miss the latest episode of The Daily Carnage Show, “The Future of Google Search,” to learn more about Google’s plans to replace the search box with 24/7 AI-powered information agents and interactive “Generative UI” that builds mini-apps and custom widgets directly on your results page.

What does that mean for advertising, for SEO, for the Internet as we know it?

Plus, catch up on Season 2:

• GLP-1s are Disrupting Every Sector (9:56)
• Spirit Airlines 2.0: Will it Fly? (11:08)
• Major Platform Updates (8:17)
• Nike Runs into Controversy (12:19)
• Transgressive OOH (18:45)
• Your Brain on AI (12:33)
• Meta, McDonald’s Mukbang, and The OpenAI Mess (14:41)
• Ethics in the AI Era (18:33)

🎙️ Now streaming on Youtube, Spotify, Amazon Music, and Apple Music.

“The Daily Carnage was the newsletter that brought us the most meaningful demo clicks and new trials.” — Digital Summit

“Open rate was nearly 50% and CTR was relatively high, especially considering this newsletter’s audience was likely being exposed to our event for the first time.” — Adweek

Hop on our 2026 sponsorship calendar to get your content in front of 25,000 marketers.

Book a $99 Classified ad instantly, or ask us about our custom sponsorship packages tailored to your campaign objectives.

Squirrelfriends, check out this campaign by Canva.

In a stunt created with Quality Meats, Canva created a fictional “squirrel worship” movement in NYC by planting a statue, launching “Squirrelites” experiences, and seeding social and influencer content.

Yesterday, they revealed the punchline: this is what happens when even the strangest idea is given the tools to look real.

Weigh in on The Daily Carnage Subreddit:

• The Daily Carnage Show Ep 2.9 Discussion: “If AI agents become the primary interface to knowledge, what happens to the humans and businesses that knowledge originally came from? What does it mean for the internet to quietly transition from an open ecosystem of destinations into a closed ecosystem of gate-kept answers? Are we speeding toward a future where people stop visiting the web directly? If that behavior ends, we ever be able to rebuild it?”

General Electric, 1950s