---

Everybody has a two-dollar bill in their pocket right? No? Just my Grandpa who has three of them? Well, if you have a two-dollar bill you’ll see a painting by John Trumbull; Declaration of Independence. Forty-two men in a warm, wood-paneled room in Philly, drafting a country into existence.

During the Fourth of July I came across a tweet and it got me thinking about the scene and the fellas who were there. “What would that same scene look like in 2026?” Most of those guys in ‘76 were in their late twenties / early thirties. So who’s in the chairs today?

Honestly, though, I struggled to come up with an easy answer. Plenty of early stage tech founders are outsizedly impactful, for sure. But name any twenty-six year olds who are being handed the authority to redesign the operating system of a republic. We don’t have the same raw influence today. Plenty of brilliant young people, but none of them are being trusted with civilizational-scale design authority. That “room where it happened” doesn’t really exist in the same way.

We didn’t lose those kinds of impactful rooms because we ran out of geniuses, but because we passed laws against the conditions that make them possible. It wasn’t malicious or even really on purpose, we did it and continue to do it because we still don’t appreciate the implications of it.

Powdered White

One important caveat to my thinking. The Founders weren’t a bunch of random kids. It was more like a barbell. On the one end, you had elders bringing legitimacy to the table. Franklin was 70, Samual Adams was almost 54, George Washington was mid-40s. Then, on the other end, you had a bunch of twenty-somethings providing energy. On July 4th, 1776, James Madison was 25, Alexander Hamilton was 21, James Monroe was 18. David McCullough has a good line about why we get this wrong:

“We tend to see them as much older than they were,” he says, because we know them from the portraits painted decades later, “when their hair, if it hadn’t turned white, was powdered white.”

We’re all familiar with [insert monument here] but so much of the work that earned them the marble was done, basically, as kids.

The thing that vanished wasn’t “young people doing things.” Young people do plenty of things. Some of them sick, some of them (as the youths would say) mid. But that barbell of a room where a 21-year-old holds real genuine authority, and a 70-year-old stands behind him pointing out the potential mistakes? That doesn’t exist anywhere.

Walter Isaacson, in The Greatest Sentence Ever Written, draws a line between the two founding instincts on this. Jefferson wanted a “natural aristocracy,” with things like a school system where “the best geniuses [would be] selected,” so that “twenty of the best geniuses will be raked from the rubbish annually.” Raked? Gross.

Franklin, as Isaacson tells it, “correctly saw the danger of creating a meritocratic aristocracy,” and built what became Penn to enrich everybody rather than filter an elite, aiming at what he called “true merit,” which he defined as “an inclination joined with an ability to serve mankind.”

Elders who want empowerment and youths who want power. It became a pretty powerful culmination of forces.

Talent Tracks Leverage

Okay, so strip the founding down to the core and you get a small, dense, canon-sharing network of young people converging on a single high-leverage project at the exact moment the institutions in that domain are still moldable. That describes the Founding Fathers. It also describes the traitorous eight walking out of Shockley Semiconductor, the Manhattan Project, Bell Labs, the Vienna Circle, etc. Those kinds of energies are episodic, often defining a pretty critical generation. Talent tracks to leverage; give a group of sharp people the ability to actually influence outcomes and they often do.

I’ve written before about this, in the context of what I called talent vortexes: very talented people want to associate with other very talented people, and the association becomes self-perpetuating. The cleanest modern example is stuff like the PayPal Mafia. Thiel, Musk, Hoffman, Levchin, Sacks, Botha. Late twenties and early thirties at the time. They dispersed and seeded SpaceX, LinkedIn, YouTube, Palantir, Founders Fund. That can feel like a super different analogue to the Founding Fathers, but that framework of a group that stayed a network and kept re-founding through each other’s influence feels a lot like the Virginia Dynasty that occupied multiple presidencies and pivotal Founding decisions for thirty years.

Today the whole American Dynamism cluster around the Gundo, Anduril, etc. is revolving around an explicit “re-found American state capacity” language, which is kind of self-consciously grabbing for founding-era vibes. And then there’s the AI labs, which are structurally even closer to the Bell Labs energy where you have a tiny number of (often) young people who genuinely believe they are making the most consequential design decisions in human history, sharing a dense internal canon, convening physically, treating the moment as similarly moldable. Despite plenty of hucksterism, its very close to a Founding Cohort structurally.

One of the things that did strike me though, is that the civic version of this is probably dead. It’s borderline impossible for this kind of energy in politics. Maybe you’d point to AOC or Mamdani, but those people bum me out so I’ll just hope we’re not gonna get a Communist ReFounding anytime in my lifetime. But outside the civic arena, the pattern still exists.

In 1776 the highest-leverage design problem on Earth was political institutions, so that’s where the ambitious 25-year-olds went. Today the highest-leverage problems are technological and industrial, so that’s where they go. The result? We now produce founders of companies, not founders of a republic. We get cap tables instead of constitutions.

Couple of observations as I reflect on them. First, institutional density. Founding is typically a young person’s game only when there’s nothing there yet. American politics has matured and gotten choked by incumbent politicians and ideas with massive seniority-focused gating mechanisms so that you can’t really touch design authority until, at least, 55. Second, scale. One person’s leverage in a 2.5M-person colonial society was huge, and pretty malleable; just write an influential pamphlet and boom. Halls of power. But in a 340M-person continental empire, that same leverage is like… zero. So ambition gets routed differently. The Founders of America got to be young cause the institutions were empty, but now the opposite is true.

So What’s the Modern Bell Labs?

Okay, so if we can’t rebuild the Continental Congress, what can you do? Maybe try a different framework that’s more modern, post the hardening of American political institutions. What about this one: where’s the modern Bell Labs?

Not another AI lab. Not another defense-tech startup. Something that mixes the young and the old in a genuinely Herculean intellectual effort. I hear plenty of “let’s rebuild Bell Labs” vibes but they usually are focused on copying the aesthetics. Smart people, pocket protectors, consolidated campus, intellectual freedom, etc.

Instead, there needs to be intellectual parity instead of aesthetic similarity. There’s a great term the Bell engineers used that I love: steam engine time. The idea, in Gertner’s telling, is that the knowledge required for an invention accumulates until the invention becomes basically inevitable and several people are poised to make it at once. James Watt didn’t conjure the steam engine out of nothing. By the late 1700s the pieces were sitting there, and he was the natural next step in assembling it, not just an isolated stroke of genius.

My whole argument is that the Rooms where stuff happens have a steam engine time too. Institutions show up when the right conditions coalesce, and they’re dormant when the conditions aren’t there. Bell Labs’ intellectual parity doesn’t really show up these days not because there aren’t geniuses, but because the structure to support open-ended research was basically outlawed in private (e.g. non-governmental) markets.

So if we want the intellectual vibes back it can’t just be “hire geniuses and give them a nice building.” There has to be an attempt to reconstruct the structure synthetically. Shooting from the hip, there were like… three rules, maybe, that made up that structure in my mind.

Rule #1: Pay The Rent

First, some facts. Bell Labs ran on that sweet, sweet AT&T monopoly money. A guaranteed, enormous, competition-insulated revenue stream, of which research was a small, stable fraction. As Gertner puts it in The Idea Factory:

“AT&T maintained its monopoly at the government’s pleasure, and with the understanding that its scientific work was in the public’s interest. An audacious move to capitalize on the transistor, should it turn out to be hugely valuable, could well invite government regulators to reexamine the company’s civic-mindedness and antitrust status.”

Bell Labs was cautiously avoiding squeezing their homegrown Golden Goose for all it was worth.

Then the 1956 consent decree made the “easy does it” mentality official. AT&T was forced to license its patents essentially royalty-free for everything it had already invented, to anyone who asked. Around 7.8K patents in 266 technology classes, were scattered to the wind. Interestingly, economists now actually credit that decree with spurring the innovation explosion that followed, precisely because it told the Labs they’d never fully own the fruit of their science. So they might as well chase the science for its own sake.

That’s a really interesting and critical takeaway that is dramatically ignored in today’s “get rich quick” economy pursuing the “creation of the perpetual underclass.” Patient capital was not a virtue anyone chose at Bell Labs, it was a byproduct of highly-regulated monopoly. In the very exceptional book, Boom, Byrne Hobart and Tobias Huber put the mechanism even more bluntly than Gertner does:

“AT&T was essentially a government-backed theoretical research program with a slight tilt toward telecom applications, wrapped in a nominally private-sector business… The result was a single company that gave birth to multiple trillion-dollar industries.”

Hobart and Huber note it was “far better for AT&T to report lower profits and spend a lot on R&D than to earn as much as its monopoly position would allow.” The research was, in part, a place to park profits the monopoly wasn’t allowed to keep. Arbitrage!

They had a business so insulated it could measure progress, in the words of Theodore Vail’s strategy, “in decades instead of years.” I’ve written before about that exact line, in a piece about how badly humans plan in ten-year units, and about how centuries are a reality for businesses the way that decades are a reality for people. Remove the “rent” and you get quarterly capital, and quarterly capital cannot fund a fifteen-year bet on solid-state physics. In the words of my handy man when I ask him to build a new shower for less than $5K; “the math just doesn’t pencil.”

I’ve argued that a moat is better understood as an engine than a wall: “a wall usually breaks down when other people try to pull them down; otherwise, they could just sit there; monopoly style. But an engine breaks down in the process of running. It requires significant maintenance.” But AT&T really did just have a wall that kind of sat there. And the weird, uncomfortable reality of a wall that just sits there is that everything behind it gets to be patient. I don’t love that conclusion, but I think there’s something to it.

This is also exactly why an AI lab isn’t the answer, despite having most of the other ingredients that feel Bell Labs-esque. Gertner nails the difference:

“Such companies don’t exist as part of a highly regulated national public trust. They exist as part of our international capital markets. … new scientific knowledge matters far less to them than the demands, for leadership, growth, and profits, of their customers, employees, and shareholders.”

The capital in and around the AI labs is speculative and competitive, not rent. It will eventually demand returns, and the research will turn inward toward product. Even Google, with their monster cash flow, was ultimately a profit-maximizing eyeball beacon. And it’s why a defense-tech neo prime via startup isn’t the answer either: a VC-backed firm gets a 7-to-10-year permission slip with an exit obligation stapled to it, which makes bottomless basic research a cost center rather than the mission.

I’ve written before about my fear that a generation of VCs are wading into a very physical, very capital-heavy world carrying philosophies that were never built for it. This is that fear at its logical extreme. As one Kleiner Perkins partner told Gertner, with a bluntness that should haunt every “rebuild Bell Labs” thread on the internet: “We don’t fund science experiments.”

Granted, the AI labs and the defense neo primes are the closest things we have, and I’m no hater. I’ve funded plenty of would-be and have-become defense neo primes, and at least one or two AI labs. They’re structurally near the ideal, and for that we love ‘em to death. But they don’t meet the bar on Rule #1.

Rule #2: Mine Bottomless Pits

Bell Labs’ charter was not “do great science.” Pure “advance human knowledge” is a recipe for random science fair drift. But pure product shops go shallow; monetization slop as it were. The Bell Labs mission was narrower and stranger than either: transmit information reliably across a continent. Mervin Kelly described the phone system as “a single, integrated, highly technical machine” connecting “more than 40 million points to any one of all the others.”

That was a honest-to-God engineering problem with a shippable product. But pursue it honestly and it immediately takes the shape of countless physics and math problems. You want a better switch, and you get the transistor. You want to know the actual limit of a communications channel, and you get information theory. Claude Shannon didn’t set out to invent the information age. Like a boss, this is how he talked about his own work:

“I am very seldom interested in applications. I am more interested in the elegance of a problem. Is it a good problem, an interesting problem?”

Bell Labs just happened to be a place where the good problems and the phone network were the same problems. Which is the thing Gertner keeps circling: the Labs had “plenty of good ideas out there, almost too many. Mainly, they were looking for good problems.” Chefs kiss.

That’s the shape you need. It’s neither “advance knowledge for the sake of progress,” nor is it “ship a product people want.” (I’m looking at you Paul / Sama / Garry). The goal is a concrete deliverable whose honest pursuit turns out to be inexhaustible. The mission was to find bottomless pits that kept unearthing new deposits to mine.

Rule #3: Have a Loading Dock

One critical additional point of context. Bell Labs was welded to Western Electric, its manufacturing arm, and to the operating companies, all serving a continental install base. This is a rule that any beautiful think tank is always missing.

Shockley could walk down the hall to someone who could actually fabricate whatever was in his crazy, eugenics-fueled imagination. Kelly cared about this to the point of obsession. “Physical proximity, in Kelly’s view, was everything,” and he built “branch laboratories at Western Electric factories” just so his scientists could follow an invention into the machinery that made it real. Inventions got deployed at nation scale, and deployment fed real problems back up the pipe. The Labs’ own development chief, Jack Morton, had a definition of innovation that any modern research institute should tattoo somewhere visible: if you haven’t manufactured it in quantity, you haven’t innovated, and if you haven’t found a market for it, you haven’t innovated.

I’ve written before about the same arrangement at IBM, in the story Marc Andreessen tells about the “Wild Ducks.” Out of 260K+ employees, IBM formally designated eight of them:

“Most of the employees were expected to basically follow rules… Then they had this category of people they called Wild Ducks. They were the people who could make new things. And they got to break all the rules and they got to invent new products… They reported directly to the CEO, they got whatever they needed.”

The line I keep coming back to is Pmarca’s summary: “we have almost like an aristocratic class within our company that gets to play by different rules.” Which, yes, is uncomfortable next to Franklin’s warning about meritocratic aristocracy, and I don’t think that tension goes away for anybody who likely isn’t the new gentried techno-elite. But notice what the arrangement actually is. I’m not saying IBM’s bureaucratic machine of 6K+ people would invent the next great product. It was that the machine would mass produce, market, and sell whatever the eight Ducks came up with.

The Ducks were how IBM followed Rule #2. The bureacracy machine was what made it possible to follow Rule #3. Different jobs under the same roof.

Bell Labs was never a think tank. Any think tank is sorely lacking a loading dock.

Modern Foundings

Stepping back, I think all three rules also do a good job of framing the American Founding. The wealth of the new, untapped continent very much paid the rent. An infinite democratic-shaped pursuit of an “American Dream” was the bottomless pit to mine. And the need to deliver, improve, strive, grow, etc. all became our loading dock.

Okay, so apply that to today. If you were going to actually build “the thing,” mission first, following each of the Founding Rules. What’s the modern equivalent of “transmit information across a continent”?

Just one example? Energy. “Make energy clean and nearly too cheap to meter.”

It fits Rule #3 exactly. It’s concrete and deployable, because the grid is the modern continental install base, the direct descendant of the phone network. And there’s Rule #2; it’s bottomlessly deep, because pursuing it honestly forces you down through radical rabbit holes along topics like plasma physics, novel materials and superconductors, subsurface geology and imaging, power electronics, and increasingly biology. Fusion, next-gen geothermal, advanced fission, the whole storage stack: you can feel the endless drill bit hitting countless new deposits from every angle. And you don’t have to take my word that it’s the natural heir. Back in 2009, Steven Chu told a Senate committee that “to solve the energy problem, the Department of Energy must strive to be the modern version of Bell Labs in energy research.”

Funny enough, the very people who built Bell Labs said the same thing first. In 1923, the New York Times asked Frank Jewett, soon to be the Labs’ first president, what invention the world needed most. His answer ran on the front page: “the greatest present day need is the development of some new source of cheap utilizable energy.” A hundred years later, that’s still a pretty good trunk of the tree to start with.

Now, the rent problem, which is the hard one. The move is to refuse to build a standalone lab at all. Bell Labs was never freestanding. It was fused to a cash machine. So where’s the energy piggy bank?

How do you weld the research org to a real operating business that has rent-like margins and is your deployment channel? A power-generation company, a fab, a scaled-up utility. That operating business is your Western Electric. It funds the science, and it gives the science a place to land.

Granted, the (sometimes) cleaner but rarer alternative could be a pure philanthropic endowment, the Janelia model, call it $5-10B committed that’s never reallocated. But the welded version is more robust for a boring, structural reason: the incentive to keep funding it is built into the machine, not resting on one patron’s conscience.

And notice that the operating business does the work of Rule #3 for free. The thing that fabricates your reactor, or runs your grid, or operates your fab is the same continental install base that feeds the hard problems back up the pipe. No branch laboratory required. The deployment channel and the cash machine are the same entity.

Granted, the deployment channel is where the actual bodies are buried. I’ve written before about the grid interconnect queue, where the typical project built in 2024 spent ~55 months waiting, and historically only about 19% of projects that enter ever get built. Fifty-five months is such a hard number to divide by twelve and realize it’s lots of years! And as I’ve argued elsewhere, “the same forces that stopped nuclear fission will try to stop nuclear fusion.” So the loading dock isn’t free just because it’s in high demand.

Staffing The Founding

Now you come back to the barbell of young energetic geniuses and older established guides. Bell Labs’ genius was casting for roles and temperaments, not just IQ. Three layers: elder institution-builders in their 60s, the Mervin Kelly and Vannevar Bush figures, whose job is explicitly not to do the science; a thin middle of player-coaches in their 40s; and a large young cohort, 25 to 35, handed a real problem and a decade and then left alone.

There were also three archetypes, that, while they didn’t map perfectly to age, were probably even more important: the lone-wolf theorist (the Claude Shannon), the aggressive driver who wills things into being (the Shockley), and the connector who metabolizes everyone else’s work and makes introductions across domains (the John Pierce, who Gertner beautifully calls an “instigator,” something “different from a genius, but just as uncommon”). A room of ten Shannons produces nothing that ships.

Kelly understood the deal he was offering, and he told every new hire on day one, in a line I think about constantly:

“You get paid for the seven and a half hours a day you put in here, but you get your raises and promotions on what you do in the other sixteen and a half hours.”

But the elders’ real function closes the loop with the 1776 analogue. Cover. Their job is to absorb the institutional and political heat, raise and defend the patient capital, manage the government, and personally guarantee the fifteen-year horizon against every quarterly pressure to shorten it, so that the young people never once have to feel the clock. Franklin and Washington supplying the legitimacy so Madison and Hamilton could design.

Everybody Picks a Rule

There’s a whole meta-science around engineering parts of Bell Labs. And, both encouragingly and frustratingly true, is that each is following at least one of the three rules (and usually only one of them).

• Arc Institute (Silvana Konermann and Patrick Hsu, with Patrick Collison involved) frees scientists from the grant cycle with philanthropic core funding, coupled to universities. A real crack at the mission-and-patience legs.
• Convergent Research and the Focused Research Organization model (Adam Marblestone) is purpose-built for work that’s too big for an academic lab but isn’t a product for a company. Startup-shaped, time-boxed, mission-specific.
• Speculative Technologies (Ben Reinhardt) is aiming the same energy at materials and hardware, a sort of DARPA-lab hybrid.
• Janelia (HHMI) is living proof the culture still works when you fund it right: no classes, no grading, no grant treadmill, a multibillion-dollar endowment, explicitly patterned on the Labs.

Every one of these has reconstructed a model to follow one of the rules, sometimes two. Patient capital, a real mission, a culture of freedom. But what none of them is doing is following Rule #3: fusion to a rent-generating operating business at continental deployment scale; finding a loading dock.

Except.

There’s probably one organization on Earth that is following all three rules right now, though I’m not sure I love my conclusion. TSMC. Rent-like margins from a position no competitor can assault. A mission that is concrete and bottomless in exactly the Rule #2 way, because “make the next node work” drops you through the floor into materials science and optics and fluid dynamics forever. And it is the loading dock. There is no gap between the research and the fab, because the research is the fab.

Reading Chip War, I scrawled a note in the margin that I’ve been chewing on ever since: monopolies might actually be important to cutting edge technological progress. Led me right to Byrne Hobart’s Boom. And the founding bargain Morris Chang struck is the thing that makes it hold together, per Chris Miller: Chang “promised never to design chips, only to build them. TSMC didn’t compete with its customers; it succeeded if they did.”

So the three rules aren’t impossible; TSMC sort of proves that. They’re just not really all being followed by anyone in the US. And that comes back to us, effectively, making Bell Labs equivalents illegal.

We Made Bell Labs Illegal

The reason no organization in the US can really follow all three rules is that the third one, a monopoly-scale, regulation-insulated cash machine bolted to a nation-scale deployment channel, is precisely the thing that modern antitrust, capital markets, and political economy are structured to prevent.

We built the entire apparatus of the last fifty years to make sure another AT&T could never exist. Despite it being a pretty freakin’ awesome organization in its hey-day.

Granted, we probably need to regulate monopolies. A monopoly that can fund a fifteen-year physics program probably becomes the same monopoly that can strangle a competitor, set your price arbitrarily, and tell you what phone you’re allowed to plug into your own wall like a Politburo’s centrally-planned dream come true.

Bell Labs executives’ had this hubris Gertner talks about in his book; an AT&T man waving off a competitive threat with “you don’t have to worry about this, because we have the network. No one else has the network.” Gertner’s dry follow-up: “They didn’t realize at the time that anyone could build a network.” That’s the attitude the antitrust apparatus existed to break, and breaking it was probably good.

But the uncomfortable possibility is that Bell Labs was a one-time artifact of a specific, and now-illegal, industrial arrangement. We didn’t just break the attitude. We broke the balance sheet that came with it. The thing that produced the transistor may be the same thing we passed laws to kill. Now, it feels like we’ve been doing the same type of slaughter for over 50 years without ever stepping back to wonder if that’s how we should handle these kinds of forces.

The tough part is there’s no counterfactual. No control group America where we let AT&T keep their wall up and see what the 1980s would’ve ended up looking like. The monopoloy might have been bad, but it was also the only balance sheet on the continent that could really measure progress in decades instead of years.

And it isn’t just an anti-antitrust rant. The rent, by design, is gone under corporate taxation, the loading dock is buried under 55 months of interconnection queues (as just one example). And the patient capital is, increasingly, being replaced by a 10-year fund clock with an exit obligation whether you like it or not. All those capability-eroding forces make up a Rat’s Nest, and it’s the exact kind I keep writing about: not an accident of complexity, but something we built on purpose and then forgot we built.

Therefore, What?

So I come back to this question of the “modern Founding Fathers.” And I keep finding myself thinking the same thought. There are plenty of brilliant 26-year olds. Instead, the limitation is, I think, structural. Can anyone synthesize the rent without the monopoly?

Call it “Synthetic Rent.” Can a large enough philanthropic endowment, a genuinely patient sovereign-style fund, or a long-horizon operating business with rent-like margins stand in for the thing that made the original possible? What you’re really reaching for is the capabilities of a monopolist without the predatory attitude.

And notice what kind of question that is. It’s a question about capital structure and political will. It is not a question about genius. Rent, not genius. In my head, I keep staging this debate as though I’m waiting for a smarter generation to show up, but I’m reminded every day that those people are all around us. One of the most debilitating things we’re missing is more about the balance sheet with (at least) a 15-20 year memory.

One important caveat: the biggest successes of the corporate lab era “were not captured by the companies themselves.” AT&T and Xerox inventions “indirectly created trillions of dollars in market value, but they didn’t generate outsize rewards for their shareholders.”

So Synthetic Rent might be structurally unbuildable. Nobody with a fiduciary duty should rationally want it. You’re asking someone to fund the discovery and then watch the world capture it. AT&T only agreed to that deal because a judge made them, and because their monopoly meant the giveaway cost them nothing they were allowed to keep anyway. Take away the coercion AND the insulation, and what exactly is the pitch?

When I wrote about venture capital’s Fourth Turning, I described what’s happening right now as a seizing of the old guard: VCs moving directly into government roles, stablecoins pushing into monetary control, defense tech reaching for the state’s monopoly on organized violence, tech-enabled super-PACs operating as open kingmakers, network states and charter cities and the ambition to literally reincarnate new nation states. And I closed that thought with a sentence I stand by and find unsettling: we’re not making the world a better place, we’re reshaping the world in our image.

So the cohorts ARE jumping back into the civic register; that’s certainly new. But despite whatever optimism I want to have, it doesn’t seem like they’re doing what Hamilton and Madison did. Rather than sitting down to draft the commons that everyone can live by, they’re building an exit. Charter cities, network states, private capability substituting for public capacity. A lot of those things are founding energy-adjacent, but they’re all pointed away from the republic rather than into it.

The Founding Father’s real invention was more than just an institution. I’ve written before that “the Constitution established a framework for self-correction through protest, reform and amendments. And it worked over and over again.” They didn’t build a machine that was right. They built a machine that could be MADE right, later, by people who weren’t in the room. That is a very different act from exit; it’s also way harder. It’s also way more rare today because no one gets rich doing that.

The Long Now Foundation likes to quote an old proverb that I’ve used before and can’t improve on: “If you’re planning for a year, plant rice. If you’re planning for a decade, plant trees. If you’re planning for a century, plant people.”

So go find a monopoly cash-flow to pay the rent. Dig in bottomless pits exposing infinitely fertile wells. And have a loading dock that keeps you anchored to the iterative feedback cycles of the real world.