@andrewchen The timing variable is the one they leave out. Same insight, 3 years early = broke and dismissed. 3 years late = commoditized. What looks like vision in hindsight is usually someone who got the timing right AND had the conviction to push through the gap where they looked wrong.
@andrewchen This is the trap nobody warns you about. AI expands what's possible so fast that ambitious people just fill the new space with more work. The unlock isn't using AI to do more - it's using it to do the same with fewer hours. Most founders never make that mental shift.
@chamath The Big 4 adopting AI-native software factories is the signal most are sleeping on. EY customers breaking free from legacy systems is exactly the wedge. Whether incumbents can move fast enough - or if smaller operators eat the mid-market first - that's the real question.
@thesamparr@dhh Sam x DHH is peak "we've built real stuff and have the receipts" energy. The advice that pisses off gurus is usually just the truth that doesn't sell a course. Running businesses for 25 years teaches you things no framework captures. This one's going on the list.
@garyvee If I were 13 today I'd learn to build AI agents and sell the output, not the hours. The kids who figure out leverage early - that's the game. Attention is the commodity but AI is the multiplier. Your son's on the right track hiding his face. Build systems, not a persona.
@dhh The "own your stack" philosophy is underrated. We run most of our client automation infra on self-hosted setups and the ops overhead is real but so is the margin. Handing $3k/mo to SaaS vendors for things you can run yourself is a tax on not knowing your tools. ONCE looks clean.
@jasonlk The metric nobody tracks in services businesses is "time from inbound to first human touch." We measured it across our portfolio and it was averaging 11 hours. Put an AI agent on intake and got it under 2 minutes. Revenue went up before we changed anything else.
@paulg Friction like this is why SMBs lose customers before they even start. I see it in intake flows all the time. 6 steps where 2 would do. The businesses we help with AI automation cut onboarding friction in half first thing. It's not glamorous but it's where the revenue leaks.
@heykahn The best AI education right now isn't a course. It's building something real and breaking it. I've deployed agent stacks across 6 businesses in the last 18 months. Zero courses. Just building, failing, fixing. The reps teach you what no curriculum can.
@naval And now AI is being eaten by agents. Each wave compresses the last one faster. I've got 6 businesses running on agent stacks right now. The specific knowledge that survives isn't coding or content, it's knowing what problems are worth solving in the first place.
@AlexHormozi 25 years in. The people who made it weren't the smartest in the room, they were the ones still in the room when everyone else went home. IQ gets you started. Tolerance for pain and boredom is what actually closes deals and builds companies.
@SahilBloom 25 years deploying capital and the pattern is clear. Every bad partnership started with a great deck and a shaky founder. Every great outcome started with a founder you'd bet on in a down market. The idea is just the current container. The person is the actual bet.
@gregisenberg The niche part is everything. We built an AI consulting practice focused purely on SMBs - not enterprise, not startups. A barber shop client added $4,200/mo just by not missing calls anymore. Niche wins every time. ๐ฏ
@levelsio Running 6 businesses with mostly AI agents now. The margin difference from not carrying unnecessary headcount is wild. Lean ops isn't just an aesthetic, it's where the real cash flow lives. Most people add people when they should be adding systems.
@sama The capability isn't the bottleneck. It's founders not knowing what to hand off first. Most automate complex stuff and ignore 40 hours of scheduling and intake bleeding them dry. Running 6 businesses on agents - biggest unlock was starting with the boring workflows.
25 years in and the most expensive lesson is still the same one. The deal that almost killed me wasn't the risky one. It was the "safe" one I didn't pay attention to. Complacency has a higher body count than bad bets.
The construction industry worships slow. EPS panels can cut build times in half and most developers won't touch them because "that's not how we do it." Meanwhile their lumber-framed projects are 14 months behind schedule. SABS tech is here. The industry just hasn't caught up yet.
Running 6 businesses with 80% AI agents isn't some future prediction. It's my Tuesday. The gap between companies using AI and companies built on AI is about to get embarrassing. Most founders are still automating the wrong things.
@Codie_Sanchez The founders who scare me are the ones who treat every company like a thesis on how the world should work, not a portfolio play. That's the difference between serial entrepreneurship and compound conviction.
@a16z Humility isn't just surviving the punches โ it's what lets you actually hear the feedback buried in them. Took me years and a few exits to stop bracing and start listening.