Markets rarely shift just because a new model appears. They shift when the culture on the ground is ready. Most places don’t need another ChatGPT anyway. Some markets welcome a push, others barely tolerate fast change. Being essential at home beats being "innovative" in the abstract.

Every month feels different, but December hits in its own way. Something in it brings back that small, familiar spark we had as kids. Enjoy the weekend.

These curves make one thing obvious: we're not in the old venture world anymore. Three companies pull ARR like this, and the rest of the market just becomes their orbit.

"Team" still matters, but an early-stage startup has its own texture. Most first-year teams lean on founders who jump between sales, product, and support in the same day. Experience helps, but the ones who break through are usually the ones who move faster than the market and stay steady.

Not every experiment deserves attention. We chase quantity because it looks like movement, but the meaningful tests are the few that actually change the direction you’re on. Everything else easy to undo.

When I was younger, Friday bars felt like a mandatory step in any career. Today many founders skip alcohol and late nights and choose sleep, training and wellness instead. Wellness is turning into a new form of social life where people connect through movement, recovery and calm 👍.

There are two content strategies I keep seeing. One says write only when you have something to say. The other insists a real founder always has something to say. The truth is somewhere in the middle. Some days you share insight and some days you just hope the algorithm forgives you.

I have met founders who only want the investor's money. No interest in a real dialogue and no interest in learning from anyone else’s mistakes. That is fine. Some people only grow through their own experience. But not with my capital. There are better matches for both sides.

This point resonates. I'm mostly focused on the early stages - the part of the stack where value is built with product, customers, and iteration, not financial engineering. Liquidity cycles still matter, of course, but their impact is far more severe at the top than at the beginning of the journey.

Many B2B startups chase "big company pains" that no one really owns. As Sara Santanen (FletchPMM) notes: if no one feels the pain personally, no one budgets for the fix. Focus on problems with a clear owner - that's your real sweet spot.

For me, the real startup spirit is a rabbit: it multiplies the way a startup should iterate - fast and often. You launch a version, it flops, you launch the next. Big companies win with resources. Startups win by out-breeding them with ideas.

Small home markets quietly train founders to run tight and efficient. Every dollar and every feature has to earn its place. When these teams finally step into bigger markets, their unit economics can look almost unfair.

The most expensive mistake in VC is treating social proof as proof of value. A crowd can be impressive and still be wrong. Parties, panels and press create a noisy dataset. Real signals come from customers who return, numbers that keep moving and founders who solve real problems.

In emerging markets, the real moat is the ability to operate in chaos. Tech helps, but endurance matters more. Most competitors quit long before they run out of features - they run out of patience, discipline, or clarity. But those who survive are worth watching.

Tested Threads for a bit. It feels faster than Bluesky, but there’s a lot of noise, the algorithm takes ages to figure you out, and there are fewer ways to control your feed. Funny thing: Kazakh users are unusually active - people joke it’s a local network now.

I see yet another angle here. When marketing turns into a list of tricks, it rarely works. But when you talk about what you build because you genuinely want to share it, people notice. Honest intent is easy to read. It does not replace solid execution, but it makes everything else clearer.

How’s everyone liking ChatGPT 5.1 so far? Feels like a direct answer to all the "too dry, too formal" complaints about the original 5.0. They’re pitching it as "warmer, smarter, more precise - better at following your instructions and matching your tone." Does it feel that way to you?

"The more AI we have, the more we’ll need the human kind". Can’t recall who said that. Might’ve been me 😁 .

This year is ending with unusual feelings for me. I’m usually excited about the long holidays, meetings, and trips. But this year it feels different.

Ever read a post by a famous expert and thought: "Brilliant logic, hard to argue" - only to realize later the facts were solid but the interpretations shaky? Even the strongest rhetoric is just a hypothesis if it rests on sand, not stone.

Why no “end at 30:00” button? Zoom, Meet, Teams - timers, warnings, but no auto-end. Just me, or does Big Video Call want our meetings to run long?

Charisma can open the door, but it won’t keep you in the room. Early-stage investors often fall for the founder, not the product. Yet what starts as dazzling energy can turn into stubbornness. Real leadership is proving worthy of the trust you inspired.

In every story, the main character is the storyteller. Even when it seems to be about someone else, it's the storyteller who quietly shapes everything - the plot, the focus, the cast.

YC just put out a call for startups to retrain workers for the AI economy. Replacing instead of reforming sounds bold. But maybe that’s exactly what the system needs. YC’s move adds welcome diversity to a very stale space.

Really curious to see how U.S. regulation handles this shift. Other jurisdictions have been more cautious. Let’s see if this turns into a model - or a cautionary tale.

Exactly. Innovation is not just about technology - it’s also about the rules that govern it. You can’t look at a new world through old lenses.

GPT-5 feels like a kid coming back home from college. Smarter in some ways, better at a few things… but somehow forgot half the stuff you taught him before he left.

If we're going to build gods, better a council than a monarch. Polytheistic AGI sounds like healthy digital pluralism. www.linkedin.com/posts/a16z_b...

Big VC rounds vs small ones. Sometimes moving in small steps gets you there faster, safer and, in the end, further.

How can AI help untrain your bias? According to numerous accounts - backed by research - the accuracy of AI-generated answers often depends on how much a prompt nudges the model toward the outcome the user expects. The less bias in your question, the more truth in the answer.

When you manage to create a viral post, it does stroke the ego, doesn’t it? The key in those moments is to remember: the attention of 10 right people matters more than 10,000 random ones. Especially if you're building something deep and lasting - not just riding a wave.

The bitterest disappointment comes from mistakes made in a state of absolute certainty.

There’s nothing so reliable that a curious mind can’t break it.

Who’s your Peter in investing? I don’t want weather reports from someone who just watched the forecast. I want data from the one standing in the storm. In startups, too, signal degrades with each retelling. Find your Peter. Or become one.

AI startups come in three flavors: 1. ChatGPT wrappers look slick but vanish when OpenAI sneezes 2. Vertical plays wire AI into health, finance, logistics, law 3. Core infra teams build the engines that power us all Use my checklist to see if your bet survives model churn. buff.ly/wR6IZe7

If you’re not sure what to do next - start with recovery. Rest, reset, breathe. It’s never a waste of time. Clarity loves a rested mind, and momentum returns faster when you’re not running on fumes. Recovery is doing something. It’s doing what matters.

In a world where AI answers anything, your real edge is asking the right question. “Would you buy this?” sounds smart - until 9 of 10 say yes… and none pay. Try: “How do you solve this today?” Clarity doesn’t come from volume. It comes from precision. Ask better, build better.

The skill that changed the game for me wasn’t part of any plan. I just picked it up along the way, almost by accident. Funny how the things you never aimed for can end up doing the most. Everyone has one of those. What’s yours?

The most underrated reason startups fail? Bad timing. Bill Gross says timing matters more than team or idea. Marc Andreessen calls it catching a platform shift. Today’s shift is real - LLMs, agents, economic pressure. If your timing is right, don’t hesitate. Windows close fast.

You don’t have to build from scratch. When you’re testing a new idea, speed and clarity matter more than code. A white-label product might be all you need to see if the idea lives or fades. Save your energy for what’s real.

To speak or not to speak? You pass on a startup. Something feels off - market, funnel, team, it doesn’t matter. Honest feedback might help or feel like a hit to something fragile. Still, I wonder: is it kinder to stay quiet, or to speak gently and risk being misunderstood?

MIT says heavy AI use dulls the brain. But a Vienna meta-analysis shows we're not losing focus—we're adapting. Our attention is becoming faster, more selective, more strategic. It’s not decline, it’s evolution. That too is intelligence.

🎲 Some games aren't rigged. They're just not yours. You can train harder, network wider, hustle smarter—but if the game's rules reward something you don't play for, you're just leveling up in the wrong arena. Know the game. Or find a better table.

You're not lazy. You just don’t have users yet. The moment someone actually needs what you built, motivation shows up with coffee in hand. Nothing lights a fire like a real person saying, “This helped.” Suddenly, you’re building again.

AI keeps expanding. But the human factor still decides the pace. Stanford found most startups automate tasks people prefer to keep. Real traction comes where tech aligns with trust and willingness. Build where you're needed — not just where you're ready.

Robots here, robots there — but where are you? Morgan Stanley says we'll have 1B humanoids by 2050. Most won’t clean your kitchen — they’ll run warehouses, sort boxes, and never ask for a raise. It’s a $4.7T industry in prelaunch mode.

Choosing a cofounder is like getting married. Except without the romance. You can fix bugs. You can’t fix broken trust. Test the partnership before the pitch deck. It’s not about chemistry. It’s about resilience when things go sideways.

Pretty sure I’m missing most of my contacts’ posts. When will AI step in and surface what actually matters to me? Hashtags and old-school feeds feel outdated. Time for smarter discovery.

Big Brother, lite edition. We fund tools that read faces, track clicks, and score “loyalty.” It's all SaaS and productivity—until it’s not. The moment control becomes more profitable than trust, the systems are ready. And guess who built them.

Hacking ≠ Cheating Using the system’s flaws to your advantage isn’t cheating — it’s surviving. The game was never neutral. Learning the code just means you stop playing in the dark.