Axios C-Suite: Why Husk is an influencer you should know
Add Axios as your preferred source to
see more of our stories on Google.
Every day, a guy known as Husk asks ChatGPT to do something simple — like time a mile run or spell "strawberry" — and posts the footage of it failing.
Why you should care: He's racked up over 15M TikTok likes making AI look like a confidently wrong intern, and the comment sections tell you exactly what's forming in the minds of your workforce.
- The general vibe: "Every time I start to believe humans will be replaced by AI, one of your videos pops up and I'm instantly returned to reality."
Between the lines: The videos aren't wrong. The failures that Husk documents are real.
- OpenAI CEO Sam Altman confirmed that ChatGPT's voice model can't start a timer on the Mostly Human podcast when confronted with a Husk video earlier this month, calling it a "known issue" that would take a year to fix.
- Husk's audience draws the broadest possible conclusion from the narrowest possible failure. A chatbot that can't run a timer becomes evidence that AI can't do anything useful at all — an inference that's costing you adoption inside your own walls.
By the numbers: Look past Husk's follower count. That matters less in 2026. The algorithms love him because people engage with his content in droves.
- The average engagement rate (likes plus comments plus shares divided by followers) on Instagram for accounts of Husk's size is under 1%, per PostEverywhere. Husk's posts consistently crack 10%. He's had multiple recent posts in the 35%-55% range.
- His video featuring ChatGPT reacting to Altman reacting to his video (a lot to keep up with) is hovering around a staggering 225% interaction rate.
The bottom line: Content that makes AI look stupid spreads far faster than content that shows it working — and your employees are hitting the share button.
📈 If you're a CEO or on a CEO's team: Request to join Jim's new Axios C-Suite weekly newsletter.
