Talent OverDrive!

@asmartbear Work to out-market the competitors. Hire their marketing team or find some other angles.

Your competitors are out-marketing you. That’s why they’re winning. If you can't compete on distribution--often the case, even with VC backing--what can you do? 1. Niche down 2. Creativity 3. Ask happy customers to help Any other ideas?

@hthieblot Everyone loves the idea of being a founder - until rent is due.

Everyone loves the idea of being a founder. Until you’re 1 year in, broke, have zero traction, and your family is asking when you’ll get a “real” job.

@asmartbear At some point "discovery" is just an inferential simulation. So building *is* the next phase of discovery.

At some point, you have to stop trying to do more "discovery" and just build. When? At minimum, when you stop being surprised, because that means you stopped learning. Even better: hard time-box on discovery, to prevent analysis paralysis. How else can you know?

@hthieblot relentlessness > resilience

Most founders get knocked out hard, multiple times. Products fail, endless pivots Fundraising falls apart. Co-founder quit Competitor crushes them Most people stay down after the first few KOs The ones who win keep getting back up. In the long run, relentlessness beats talent

@asmartbear *done

@asmartbear Yep, true "building in public" is "boring in public". This is why so many have this twisted view of how business gets down and who does it.

I’m so naive I thought “building in public” meant honestly sharing the ups and downs, so others can commiserate and encourage and celebrate and even give feedback. As opposed to posturing, going for “viral” posts, and sometimes just lying. Again, naive, I get it.

Major green flags in a founder: 1/ crazy grit, won’t give up 2/ started building before pitching investors 3/ constantly moving, constantly generating new information 4/ very technical 5/ not addicted to founder cosplay 6/ goes out and talks to users 7/ finds creative, experimental marketing channels 8/ genuinely cares about what they’re building 9/ never blames others or external factors 10/ tinkered with projects obsessively at a young age 11/ constantly comes up with interesting ideas 12/ knows their metrics cold 13/ doesn’t chase hype, chases truth 14/ excellent storyteller. Can sell the vision to hires, investors, and customers 15/ can clearly explain what they are building in one sentence 16/ can’t stop talking to customers If you hit a good chunk of these, tell me what you are building

@hthieblot 5/ no cosplay or larping as [Steve Jobs / insert model]

Previous list from 3 months ago: x.com/deedydas/statu…

The Ultimate List of Artificial Intelligence "Neolabs": May 2026. A Neolab is a pre-revenue scale startup working on long-term AI breakthroughs, usually with a $1B+ valuation. There are now 63 of them!

Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.

Just met an engineer at a top AI lab. Claims nobody goes to the office anymore, they just have agents running the work while they manage everything from a phone. Man is literally "working" from coffee shops and the beach. The future is wild.

Researchers reverse-engineered Anthropic’s leaked Claude Code. And what they found completely shatters how we think about AI agents. We all assumed the smartest AI tools had complex, sophisticated "brains" doing the heavy lifting. That the model itself was doing the real work. When researchers cracked open the 512,000 lines of leaked source code, they found something shocking. Only 1.6% of the codebase is actual AI decision logic. The other 98.4%? Operational infrastructure. The core AI agent is literally just a basic while (true) loop that calls the model and parses the response. That’s it. The real magic of Claude Code, the reason it feels so capable, reliable, and autonomous, has almost nothing to do with the LLM itself. It’s the harness. Anthropic built hundreds of thousands of lines of pure infrastructure just to keep the AI from hallucinating, breaking, or forgetting what it was doing. - A five-layer context compaction pipeline to manage memory. - A deny-first permission system with seven safety modes. - A secure shell sandbox. - A complex orchestration pipeline just to route tools. The implication for everyone building AI right now is massive. Stop obsessing over the model. The model is just a single function call. Even the most advanced frontier model in the world is completely useless without an incredibly robust engineering harness around it.

@hthieblot Vanity Metrics kill companies, especially startups.

As a founder you must resist all the things that make you feel successful without actually making the company successful. -raising big rounds you don’t need -Saying yes to every podcast, panel, and networking dinner -hiring a big teams without PMF -getting a big office Etc.

@marcrandolph I don't know how the goal "of being rich" is going to carry anyone to the finish line of a startup. Better to join a well funded startup as an early employee and hop into the wage cage for 4 years to fully vest.

It amazes me how often entrepreneurs tell me their number one goal is to be rich. I find this tragic. First, it’s highly unlikely. It’s rare enough to earn a living at this game, much less accumulate life-changing wealth. Second, it misses the real reason entrepreneurship is so rewarding: getting to sit around a table with smart people solving interesting problems.

Though rare, there are some small percent of founders who left to implement their vision that may have including optimizing not just the product or processes but the talent and teams. Often, the very best performers and leaders in a company are swarmed by toxic or terrified peers and knifed before being fired and discarded. Common occurrence as an employee but in rare cases such a person might found a company as well.

Founders start their own organizations in part because they’re proactively searching can’t abide being a “good team member” in someone else’s. And then you wonder why they’re usually terrible at team building and culture. Both VC-funded and bootstrapped.

@pedroldrh @hthieblot Why do schools keep teaching the Pythagorean Theorem and Quadratic Equation? I do believe these are important keys to success but people should find these on their own.

@hthieblot Why do you keep posting about this topic? It feels like you are leading a generation of founders astray. I do believe that grinding for months and years is the way to success but people should find this on their own

Everyone loves the idea of being a founder. Nobody loves the reality: 3 years in, 6 pivots, $0 in the bank, and a co-founder who quit. Parents doubt you. Investors Ghosts. But you still show up at 120%. The exit shows the world who you are. The struggle reveals who you are.

@hthieblot And if you can sell, you don't need anybody.

All you need to build a company

As the VC industry matured and funds ballooned, we also say Angel investors/investing mature and the numbers of angels and investments they would make ballooned. So to deploy their huge funds and get required returns, VC went to pour $$$$$$$ into "sure things" and the unpolished gems/uncut gems moved to the mature angel layer.

Venture capital used to be about finding the hidden gems. Now it’s entirely about backing the sure thing. Lesson if you’re raising is that if you’re a hidden gem, start to behave like a sure thing. Else you will stay hidden.