A podcast from Stupid Human Ltd
One dead startup per week.
What it promised, what it cost, what it teaches. Operator-grade postmortems — not schadenfreude — narrated like a long-form magazine essay.
Season one
Episodes
New episode every Tuesday
03
Olive AI: $852M to Discover "AI" Was a Marketing Word
Olive AI raised $852 million from Tiger Global, Sequoia, General Catalyst, Google Ventures, and Mayo Clinic. At peak, it was valued at $4 billion. By October 2023 — 27 months after that peak — the company was winding down and its 1,300 employees were gone.The thing that killed Olive wasn't competition. It was an Axios investigation that ran in August 2022, with hospital executives saying the AI they bought didn't actually do what it was sold to do. In enterprise software, where every CFO talks to every other CFO, that's a fatal event.This episode walks through the four real failures: a marketing word ("AI") that the product couldn't live up to, an enterprise market where one customer's bad reference closes every pipeline, a peak-of-cycle valuation that the cycle then ended, and the hardest kind of failure to recover from — a misrepresented product, because the gap isn't technical, it's trust.If you're building anything in enterprise SaaS, the lessons compound.—⏱ Chapters0:00 — Cold open: $852M, dead 14 months after one Axios story3:30 — Act 1: The Promise (Sean Lane, the rebrand, $800B problem)11:00 — Act 2: Why a $4B valuation didn't sound stupid in 202119:00 — Act 3: The crack (the "AI" gap, the Axios story, the layoffs)29:00 — Act 4: The wind-down and the asset sale31:30 — Act 5: What Olive actually teaches you37:00 — Outro
26 Jun 2026 · 38 min
02
Humane AI Pin: $230M to Die in 12 Months
Humane AI Pin: $230M to Die in 12 MonthsHumane raised $230 million from Sam Altman, Marc Benioff, Bob Iger, Microsoft, and OpenAI. Twelve months after launch, Marques Brownlee called the AI Pin "the worst product I've ever reviewed."By February 2025, Humane sold to HP for half what they raised. By May, the servers were dark and the pins were bricked.The founders blame timing — that the AI wasn't ready. They're wrong. This episode walks through the four real failures: a hardware-first launch when the software wasn't ready, the impossibility of replacing a smartphone by being parallel to it, an investor halo so bright nobody asked obvious questions, and the gap between what a TED demo proves and what a shipped product does.If you're building anything that depends on a category that doesn't exist yet, the lessons compound.⏱ Chapters0:00 — Cold open: $230M, dead in twelve months3:30 — Act 1: The Promise (Imran, Bethany, and the post-phone pitch)9:45 — Act 2: Why this didn't sound stupid in 202315:30 — Act 3: When the cracks appeared (the demo, the reviews, the recall)25:30 — Act 4: The HP acquisition and the shutdown29:00 — Act 5: What Humane actually teaches you36:00 — OutroAbout the showEpic Fail is a weekly solo-essay podcast that dissects one dead startup per episode. New episodes every Tuesday. Subscribe so you don't miss the next postmortem — coming up: Olive AI, the $4B healthcare AI casualty.Sources: Marques Brownlee review (April 16, 2024), The Verge launch coverage by Wesley Hill and David Pierce, The New York Times reporting by Tripp Mickle (August 2024), The Wall Street Journal acquisition coverage (February 2025), Wired and Engadget launch reviews.#Humane #AIPin #StartupFailure #BusinessPodcast
11 Jun 2026 · 38 min
01
Quibi
$1.75 billion to discover that nobody watches premium TV on the subway
13 May 2026 · 43 min
Humane AI Pin
$230M, returns exceeded sales, sold to HP for parts. The danger of building hardware before the software model exists.
Olive AI
$4B healthcare AI valuation, gone. Vibes-based valuations and the AI hype cycle's first major casualty.
Convoy
$3.8B digital freight unicorn that turned off the lights in a single weekend. Capital intensity, freight cycles, and ZIRP unwind.
Juicero
$400 Wi-Fi juicer that you could squeeze by hand. The pure hubris episode.
The format
Five beats.
28 minutes.
A consistent structure makes the show recognizable. The format itself is the brand.
- 01
Cold open
The most dramatic moment — the shutdown email, the SEC filing, the all-hands where everyone got fired on Zoom. In media res.
- 02
The promise
What they raised, from whom, the founder's pitch, the magazine covers, the valuation peak.
- 03
Why it sounded brilliant
Steelman the original idea. What problem they were solving, why smart people believed.
- 04
The crack
When things started going wrong. Visible signals vs. hidden ones. Where the unit economics, product, market, or governance broke.
- 05
The death and the lesson
The actual end, plus 3–4 takeaways framed as principles, not hindsight bias.
Listen
Subscribe wherever you listen
Or stream the latest episode right here.