🚀 The Real Story of YC at 21 Years: What Makes Founders Formidable

🎯 The Fundamentals Never Change

Y Combinator recently completed its 47th batch — marking 21 years of the world's most influential startup accelerator. Despite persistent claims that "YC has jumped the shark," the fundamentals of building successful startups remain remarkably consistent across two decades.

Paul Graham, YC's co-founder, still prepares each batch's opening talk from scratch. Not because the content changes dramatically, but because the core principles of starting a company — whether in microprocessors, internal combustion engines, or AI — stay constant. The technology changes. The ambition required does not.

"Most of starting a startup is always the same. It doesn't matter if you're working in microprocessors or AI or the internal combustion engine — it's always the same stuff."

📈 The "Supposedly Good Old Days" Weren't Actually Better

Critics have been declaring YC's decline since approximately 2008 — a narrative Graham calls the "supposedly good old days" myth. The pattern is predictable: detractors can't claim YC was always terrible, so they argue it used to be great but has now fallen off.

The reality? Today's YC companies are dramatically more ambitious than early batches. While early successes like Reddit created genuine value, current companies are tackling challenges like:

• Intercontinental ballistic cargo — literal ICBMs that land and deliver payloads instead of exploding
• On-demand cancer research — personalized treatment protocols for individual patients
• Cancer vaccines and novel therapies — attacking the disease from multiple angles simultaneously

One standout from the current batch provides customized cancer research for patients — the same team that helped GitLab CEO Sid Sijbrandij successfully navigate his own cancer diagnosis. Sid famously "went founder mode" on his cancer, treating it like a startup problem to solve. Now that approach has become a scalable business.

🔥 What Makes Founders "Formidable"

The term "formidable" comes from Graham and Jessica Livingston's private vocabulary, developed before YC even existed. The definition is deceptively simple: someone who gets what they want.

This matters immensely for investors. When a founder is formidable and an investor owns equity in their company, interests naturally align. If the founder gets what they want, the investor gets what they want. It's elegant incentive design.

"Do you get what you want? That's the test. Because how formidable are you if you don't get what you want?"

Importantly, this quality is largely inborn rather than taught. Sam Altman was already "extremely formidable" at his first YC meeting, before even being accepted. The rare exceptions involve founders who possess ambition but have been trained not to show it — typically by pushy parents or institutional environments demanding obedience.

⚡ What Actually Motivates Founders Day-to-Day

Contrary to popular belief, the prospect of becoming a billionaire isn't what drives founders through daily challenges. The real motivator? Fear of failure.

When servers crash or critical systems fail, founders aren't thinking about future wealth. They're thinking: "The engine of my model train set is falling off the table — I need to save it." It's about preventing disaster, avoiding embarrassment, and protecting what's been built.

Founders often work heads-down for years, only lifting their heads to discover — sometimes from Graham himself doing the math — that their shares make them billionaires on paper. The realization takes them by surprise.

This is also why treating YC as a "resume badge" fundamentally misunderstands what's being signed up for. A Harvard degree can be coasted through with easy majors. A startup has no easy mode — it's like being forced to study theoretical physics with no option to switch majors.

🤖 AI Changed Everything (And Nothing)

Graham studied AI in the 1980s — a vastly different field that, in his words, "would never have worked" and "was a joke." The assumption then was that AI would evolve like biology: start with a perfect fly brain, progress to mice, then cats, then monkeys, and eventually humans. Perfect execution at each level, gradually increasing in complexity.

What actually happened was the opposite quality optimization. AI arrived as fully human-level capability but initially terrible at accuracy — like "an undergraduate trying to BS his way through a paper." Instead of perfect-but-limited, we got capable-but-unreliable. A slot machine generating plausible-sounding human text.

"Nobody expected that when the first plausible AI showed up, it would be like a bullshitting undergraduate — just a slot machine spitting out words."

This created what's now called the "jagged frontier" — AI that can allegedly solve famous open problems in mathematics while simultaneously failing to answer basic questions about restaurant hours. Some capabilities are far beyond the "finish line" of AGI; others remain stubbornly behind it.

The Turing Test, once seen as a clear threshold, now reveals itself to have width rather than being a simple line. What looked from the 1980s like a sharp horizon is actually a smear — and current AI exists somewhere in the middle of that smear.

📦 Shipping Speed Still Matters — Even in the AI Era

Despite powerful AI tools that can spin up parallel workstreams and automate development tasks, the best predictor of startup success remains unchanged: the pace at which companies ship new products and features.

Many startups in the current batch still aren't shipping fast enough, proving that AI hasn't eliminated the fundamental requirement for rapid iteration. The constraint isn't purely production speed — founders must still generate ideas, make decisions, and execute strategically.

The one genuinely new variable? Massive AI bills. Startups traditionally had one major cost center: salaries. Everything else — laptops, software, infrastructure — was cheap by comparison. Now companies face tens of thousands of dollars per day in token costs, fundamentally altering startup economics during the early stages.

🎓 The YC Batch Advantage

The batch model, now core to YC's identity, was actually discovered by accident. The original concept was simply to create an "angel firm" — a VC-style operation making small, early-stage investments with standardized paperwork. The batch structure emerged as a learning mechanism: fund many startups simultaneously to quickly learn how to be investors.

The timing (summer) and duration were chosen to replace college internships at companies like Microsoft. The founders would be students who "won't mind if we're not real investors because they're not real founders." Within a single cycle, both sides became real.

The batch provides three critical advantages:

• Colleagues: Starting a company is normally isolating. The batch creates an office-like environment where founders work on separate companies but share the experience.
• Peer Learning: Technical problems are likely solved by someone else in the batch. Founders can simply ask how others handled similar challenges.
• The YC GDP: No matter what product is being built, other batch companies serve as ideal early customers — early adopters who decide quickly and must at least hear the pitch.

As YC has grown (from 40 startups in early batches to much larger cohorts today), the structure has adapted by organizing founders into pods of approximately 70 startups — recreating the intimacy of 2012-era batch sizes within a larger whole.

🔮 Where the Next Trillion-Dollar Company Comes From

The answer isn't a specific industry, technology, or market opportunity. The next trillion-dollar company comes from the right founders — formidable individuals working on promising (but mutable) ideas.

"The next trillion dollar company comes from the next trillion dollar founders. And they probably have good ideas — whatever they're working on is probably promising."

After 20 years of data, founder characteristics remain remarkably consistent. The traits that predicted success in 2005 predict success in 2025. There's no evidence this pattern will change over the next 20 years.

Unless, of course, the founders are robots.

✨ The Unchanging Core of Startup Success

In an era of exponential AI advancement, geopolitical instability, and rapidly shifting market conditions, what's striking about Y Combinator's model is its stability. The problems founders face, the qualities that predict success, and the methods that work have remained constant across 47 batches.

Ambitious, formidable founders who ship quickly and refuse to accept failure — that formula worked in 2005, works in 2025, and will likely work in 2045. The surface details change. The fundamentals endure.