Emmett Shear Masterclass: Founder of Twitch, CEO of OpenAI, Tech Entrepreneurship Advice
Emmett Shear walks Melanie Uno through the full arc of his career: a failed web-calendar startup that folded the day Google Calendar shipped, the accidental birth of Justin.tv as a 24/7 livestream of his co-founder's life, and its pivot into Twitch — which he's quick to point out was a social-media company, not a gaming one, ultimately acquired by Amazon for nearly a billion dollars.
From there the conversation turns to what actually makes founders and teams work — why the only right response to being told to quit is "screw you," how teams "gel" when everyone starts thinking in terms of "we," and the lessons Paul Graham drilled into him at Y Combinator. It closes on his three-day turn as interim CEO during the OpenAI crisis, why he left to do basic AI research, and his clear-eyed but hopeful view of where AI is heading.
“If I can convince you to give up now by telling you this, absolutely give up.”
“A gelled team is when people stop thinking about their career goals, their needs, their project, what they want to do.”
“If you're successful, it's very stable — more so than anyone who was working for some random company or even a really impressive company.”
Key takeaways
- 00:03:43His first startup was a JavaScript web calendar; when Google Calendar launched he realized he had no edge and sold it on eBay.
- 00:04:30Justin.tv began as a 24/7 livestream of Justin Kan's life, then became a platform once the founders admitted they were better at building tech than making entertainment.
- 00:05:17Twitch was "Justin.tv with a fresh coat of paint," refocused on game streaming — and was really a social-media company, not a gaming company.
- 00:18:06Teams "gel" when members stop optimizing for their own goals and start thinking in terms of what "we" need to get done.
- 00:30:00Paul Graham's "confidence trick" — acting with conviction before you feel it — is what carries founders through the chaos.
- 00:38:30He stepped into the OpenAI crisis as interim CEO to keep the company from fracturing into multiple pieces.
- 00:52:30He had intended to move into basic AI research from the day he left Twitch — and finally did.
Chapters
- 00:00:00How I Accidentally Built Twitch
- 00:02:45My First Startup Flopped HARD
- 00:05:25The Pivot That Saved Our Company
- 00:08:10The Real Power of Building with Friends
- 00:10:45The Truth About Startup Chaos
- 00:13:30Getting Told I Wasn’t Good Enough
- 00:16:15Who Should’ve Been CEO? (Hard Truth)
- 00:18:55Why Twitch Worked & Justin.TV Didn’t
- 00:21:35How to Build Something Great
- 00:24:30The Harsh Reality of Startup Life
- 00:27:05The Moment We Knew Twitch Would Win
- 00:30:00The Confidence Trick Paul Graham Taught Me
- 00:33:00How I Rewired My Brain to Learn AI
- 00:35:50Why I Walked Away from Twitch
- 00:38:30Inside the OpenAI Crisis (Before I Took Over)
- 00:41:00What I’m Building Next in AI
- 00:44:0090% of Founders Fail Because of THIS
- 00:47:30You Can't Fake the Grind (Do This Instead)
- 00:50:00Why Being Niche Made Twitch Unstoppable
- 00:52:30The Real Reason I Left Twitch for AI
- 00:55:00How I Taught Myself AI Fast
- 00:57:30They Asked Me to Be OpenAI CEO for 3 Days
- 01:00:00What Actually Happened Inside OpenAI
- 01:03:00The Scariest Truth About AI
- 01:06:30Why I Still Have Hope for Humanity
- 01:09:30Final Words: What I Want to Be Remembered For
