Key Takeaways

  • Wispr Flow drove 30x revenue growth for its voice-to-text app through paid creator campaigns while banning generative AI from customer-facing ad creative.
  • Hector Serrano argues that social feeds are saturated with synthetic video, digital avatars, and AI statics that look off and trigger immediate ad fatigue.
  • Matt at Victor (formerly Exited Growth) automates massive backend growth workflows, but still places a human operator at the final 10% endpoint.
  • Chasing production volume with AI tools creates negative returns; reducing asset output to focus on human craft cuts through feed noise faster than synthetic scale.

Wispr Flow Sells AI but Refuses to Market With It

Hector Serrano directs content for an artificial intelligence voice tool that scaled revenue 30x. Yet when Lenny Rachitsky pointed to Serrano's tweet (“The best marketers are allergic to AI”), Serrano defended the position outright. For a company building software at the frontier of artificial intelligence, shipping synthetic ads creates immediate cognitive dissonance for consumers.

“If we on the other side go out and start making AI slop ads and videos with AI influencers and I don't know AI designs and AI statics, it's probably not going to be received really well,” Serrano said. "It just looks off."

The core issue is user pattern recognition. Feeds are flooded with algorithmic scripts, synthetic voices, and mid-tier digital avatars. When users spot synthetic assets, they scroll past because cheap production signals a low-effort product. Rachitsky summarized the contrarian playbook: “So the alpha what I'm hearing here is like don't use AI for this work.” Serrano confirmed: “Yeah, for the final output especially on the creative side like what other people are going to see, I don't think AI should touch yet.”

Keep Humans at the Final Ten Percent Endpoint

Generative software tempts growth teams to automate the entire marketing pipeline from prompt to published post. Serrano points out that top performance operators draw a strict line between internal workflow automations and public-facing assets.

“Even one of the most bullish people on AI that I know, Matt who is now at Victor or exited growth, he has the most automations I've ever seen in the whole wide world,” Serrano said. “Even then, that last 10% that's the most important always has a human on the endpoint.”

Founders turn to automated video tools to publish fifty ad variations a week instead of five good ones. Serrano argues that volume for volume's sake destroys performance on paid channels. “There's so much noise that if you actually focus more on the quality and less on the quantity and less on building all these crazy automations and hitting volume for just hitting volume's sake, it gets so easy to cut through the noise.” The sustainable edge in consumer acquisition is not synthetic asset generation; it is human taste deciding what ships.

What to Do With This

Audit your marketing backlog today. If your team is running generative video or synthetic voice tools to produce client-facing ads, shut down those campaigns immediately. Shift those dollars toward recruiting real creators on performance retainers, and require a human editor to review and polish the final 10% of every asset before it goes live.