The AI Marketing BS Index

AI hype gets a scorecard and the comments go feral

TLDR: A new “AI Marketing BS Index” scores buzzword-heavy pitches and fake-sounding claims. Commenters cheer the call-out, demand harsher penalties for cherry-picked tests and rebranded tweaks, while skeptics argue the rubric oversimplifies necessary nuance—making this a loud, useful fight over how to spot hype before it wastes our time.

Move over, astrology—there’s a new scorecard for the stars of AI marketing. Inspired by John Baez’s “Crackpot Index”, the “AI Marketing BS Index” tallies points for classic hype sins: inventing things without proof, sprinkling science-y words where they don’t belong, ending paragraphs with pseudo-profound fluff, name-dropping Ivy Leagues, making unfalsifiable claims, and touting mystery "collaborations." The creator even starts every pitch at -5 points—benefit of the doubt, before the doubt wins.

But the real show is the comments. One crowd-pleaser fumes that every CEO brags “AI made us 10x faster,” yet the economy isn’t exactly cartwheeling—receipts, please. Others want stiffer penalties: “It’s not X, it’s Y” hedging? Make it 40 points, not 20. Power users pile on with new infractions: cherry-picking tests to look good, renaming tiny tweaks as “revolutionary,” and hiding system hacks behind flashy buzzwords—translation: stop dressing up minor edits as moon landings.

Then comes the pushback. A skeptic says the list misses real sins and misreads nuance—sometimes saying “not X, but Y” is just clarity, not con. Another jokes we should subtract points if a founder actually reads their own website before pasting it. Verdict? The index is part satire, part lie detector—and the crowd wants it stapled to every AI press release.

Key Points

  • The article proposes an “AI Marketing BS Index” inspired by John Baez’s “Crackpot Index.”
  • Every concept starts at -5 points, with points added for specified marketing missteps.
  • Penalties include lack of citations/specifications and misusing scientific terminology (10 points each).
  • Further penalties target hedging language, pseudo-profound statements, unwarranted claims of mirroring nature, misuse of “emergent properties,” and Ivy League name-dropping (20 points each).
  • Heavier penalties apply to unfalsifiable technical descriptions (30 points) and unverifiable research collaborations (40 points each).

Hottest takes

"every CEO claiming AI made their company 10x faster yet GDP trends not really budging" — snapcaster
"change one tiny term in GRPO and call it a completely different acronym, 50 points" — swyx
"that language pattern is often appropriate" — PaulHoule
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.