t.co/LGxhaM4TDn

t.co/ySUpmxpwFo

Karpathy's autoresearch isn't just for AI research. You can use it for business. Imagine running 700x more experiments: - Landing pages - Creatives - AEO/SEO - Pricing - E-mails - Cold outreach - Warm outreach - AP/AR - Procurement Revenues: 📈

Every org needs a world intelligence to operate at a 10x level. It might seem daunting, but it’s worth the investment. We’re already starting to see a multiplier at my company. Here’s how to get started:

The founder who 'hired a CEO and stepped back' just shut down. Here's the email he sent me. "Eric, you were right. I thought I could hire my way to freedom. 3 years and $10M in burn later, we're done." His mistake? The same one I've made. Multiple times. He believed the fairy tale that you can bring in a CEO and disappear. That somehow a hired gun will care as much as you do about your baby. Here's what actually happened: - The CEO made 'strategic pivots' that gutted their core product. - Revenue dropped 40% in 18 months. - Their best engineers left because "the culture changed." - Customer churn hit 35% when they stopped innovating. The founder watched from the sidelines, thinking he was being smart by "not micromanaging." Truth is, no hired CEO will ever match a founder's obsession. They're playing with house money while you built the house. I've tried this move 4 times. Each time ended the same way - me jumping back in to save what was left. The successful founders I know? They stay involved. They might hire CEOs, but they're still in the trenches on what matters. There's a reason Zuckerberg, Huang, and Musk haven't "stepped back." Your company needs you more than you think. What's your take - can founders ever truly step away?

t.co/feOhkEBLqi

TBPN is a 7k-live-viewer pod that sold to OpenAI for $200M because their average clip gets 257k views. Everyone's talking about the strategy. Here's the structure: 1. Format is engineered for clippability, not the other way around 3 hours/day, live, 5 days a week — massive raw inventory to mine Guest call-ins (6–8 per show) = built-in 5–10 min self-contained segments. Every guest segment is a pre-packaged clip candidate with a hook, arc, and payoff Hosts read tweets live on air → instant meme fodder + the tweet author shares the clip back (distribution loop) "High and low" aesthetic: mahogany desk + cinematic lighting + suits, BUT content is group-chat casual. Makes clips feel premium + authentic at the same time 2. The clip itself - consistent visual grammar Aspect ratio mix: native horizontal on X (where they're platform-native), 9:16 re-cut for TikTok/Shorts/Reels Opening 3 seconds: hard cut to the punchline or hot take. No intro, no ramp Captions: burned-in, large, two-color (speaker name + quote), word-level highlight Lower-third branding: TBPN bug + topic tag always visible → instant brand recognition even if muted Length: 30-90 seconds sweet spot. Guest clips run longer (2-3 min) when the take is meaty End frame: guest's handle + "@TBPN" — clip becomes an ad for itself 3. The caption/tweet wrapper Format: "[quote]" - @[guest] — always attribution-forward Second line: context (what show, what topic) — usually one sentence Never an explainer or thread. The clip does the work. Posted from @tbpn main account AND the guest usually quote-RTs → dual distribution 4. Volume + cadence ~8-15 clips per episode posted across the day (not dumped at once) Each clip tests a different hook/angle from same episode "Most popular clips of the last 2 weeks" recap posts = second-wave distribution on the winners Weekend: compilation content, "best of" threads 5. The real edge The clips aren't actually why TBPN wins - they're a symptom. The real skill is "timeline control": reading when a topic is peaking, coining phrases that spread, getting the right guest on the day a story breaks. Clips are the delivery vehicle for a bigger game of being the center-of-gravity account for tech Twitter. It seems like a no-brainer to at least try to figure out how to adapt this if you create content.

We are hiring forward deployed engineers from around the world. The talent we’ve been talking to coming from X has been incredible. Many engineers that go through our ‘beat AI’ challenge (in my profile) have been remarkable. If that’s you, take the challenge. 😉

t.co/p3aLVdC9dC

t.co/YfSuTchDgh

t.co/IGhe1Ll9AK

@PolymarketMoney The man @geoffreywoo spearheaded many of these

t.co/8ZT2rrhihr

Coaching and higher tier content ideation - Scans daily number 1 priority and top 3 things done yesterday and scores people individually, in front of everyone via Slack. Also sends a leaderboard at the end of each month to see how everyone did. - Sits in internal calls and pulls 'compelling content ideas' to be used for organic social. Oftentimes, your best ideas will come from passionate internal dialogues. - Coaches people on running L10 meetings, sales meetings, client meetings, etc. This would be annoying work that someone would have to micromanage but it's now 'new work' that an AI can handle. And the best part? No one can get offended at the AI for its feedback. 😉

t.co/08yn9cNYdO

t.co/Cbl7y6YtNA

We're looking to hire to build this vision - Head of Product - Head of Talent - Director of Sales - GM Among other roles... Link in first comment

t.co/vT7XA1dcCk

The next $1T company: services as software. Sequoia, a16z, YC all talk about it. But nobody is talking about how to do it. Here’s how: