Shyam Sankar

In this return of great power competition, it is clear that we need to continue to innovate to provide deterrence from conflict. This requires military superiority and continuous technological innovation. When Palantir was founded there was no path to working with the Defense Department. Not a difficult path, no path. And there was exactly one path to working with the Intelligence Community, In-Q-Tel.  

In-Q-Tel is a venture capital firm that invests in commercial technologies that are relevant to national security. But this headline confuses that actual value proposition. America has plenty of capital for quality companies. A start up that can’t raise money is probably not a very good startup. It is one of America’s strengths - a deep and wide venture ecosystem. What In-Q-Tel provides that is a game changer is customers. The capital invested is often de minimis to the companies even if it is extremely valuable validation. The customer contracts, clearances, and commitments that In-Q-Tel furnishes are the fuel for innovation.  

Not enough is said about the US Govt as Customer and too much has been said and overstated about the US Govt as R&D financier. It is broadly acknowledged that in the present era the USG cannot outspend industry in R&D, acknowledging that it very much did in the early cold war. But perhaps that spending was never the key enabler outside of basic research.

Chris Miller’s Chip Wars covers the history of the semiconductor industry. In 1965, Military and Space applications where 95% of the chip market. Bob Noyce, then Co-Founder/CEO of Fairchild and future Co-founder/CEO of Intel, always envisioned a broader market than the military which meant he had to manage his R&D priorities (and he was of course right: the first Integrated Circuit for consumers was used in Zenith hearing aid that was initially designed for a NASA satellite). He declined most military R&D contracts (despite these customers representing 95% of his revenue) so he could stay in control of his R&D roadmap. He never let more than 4% of this R&D budget come from Govt contracts. So 96% of his R&D was self-financed from investors and profits.  

In Noyce’s own words “..there are very few research directors anywhere in the world who are really adequate to the job at assessing Fairchild’s work and they are not often career officers.” Noyce complained about the time spent writing progress reports for the bureaucracy for any of this government funded R&D. Now Fairchild had the unique luxury of being funded by the 1950’s equivalent of a billionaire so they could treat the military as the customer for their products rather than as their boss of their R&D from day one. 

Because of the difficulty of achieving scaled procurement in US Govt it is unfortunately still true that a uniquely hard head and patient sort of mission-driven capital is still required today. SpaceX, Palantir, Anduril all have billionaire founders whose commitment to the cause was required to survive the various valleys of death and incredibly long acquisition cycles. Notably 2 of the 3 had to sue the Govt to ensure the Govt didn’t de facto compete against industry.

Even in Noyce’s era, the Pentagon was more comfortable working with big bureaucracies than nimble startups, and as a result underestimated the speed by which Fairchild would transform the industry. DoD assessment praised RCA for having the best microsystems electronics miniaturization program while dismissively noting that Fairchild only have 2 scientists working on it internally while Lockheed had 50, implying that Lockheed was far ahead. Of course it was Fairchild’s R&D team that made the breakthroughs. 

Defense contractors thought of chips as the final product. Noyce and Moore were already dreaming of computers and phones. Noyce slashed prices to access the broader vision and market. Cost plus government contracting would never have created the price performance the USG needed to create the incredible military deterrence capability. And of course all that commercial innovation led to a semiconductor revolution that created vast American prosperity, the underpinning of our national security. 

We see a very different story with Drones. General Atomics invented the modern drone in the 1990s with the Predator. A Noycian figure would have seen the vast potential for not only commercial drone applications but also the consumer market. Instead for decades that vast R&D focus of these platforms was locked by Govt R&D programs. And with great effect. But without any American prosperity that should have followed GA owning or spinning out a Commercial subsidiary that should have been the DJI of America. Instead the vacuum let DJI fill it to serve CCP civil-military fusion aims. Now the hobbyist consumer’s drone purchase funds CCP R&D against America.

The root problem is that the acquisition system is unintentionally communist (Bill Greenwalt noticed the same). Profits are capped at a very low number on contracts. While the law says the government should favor fixed priced contracts in practice these acquisitions happen cost plus. This means the only way to make more money as a contractor is to find a way to spend more money in your costs. That won’t work. Noyce made MORE money as he lowered the cost with even greater the profit margins. He had the commercial incentive to do that and that benefited every customer, including Uncle Sam. Noyce could not have done that rationally if he was a government contractor - lowering the price would have been lower profit with government regulated profit margins. The reason we lost the market with Drones is that China executed with true capitalism with DJI while, ironically, America pursued communism.

“Selling R&D to the government is like taking your venture capital and putting it into a savings account. Venturing is venturing. You want to take the risk.” - Robert Noyce

Noyce ultimately left Fairchild, a company he co-founded and ran, because the billionaire financier of Fairchild didn’t think anyone but he should have equity in the company. Ironically at the time equity for employees was viewed as creeping socialism. So the talent left. Noyce and Moore left to found Intel. Remember that Moore’s Law is not actually a law at all. It is a goal. To the extent Intel achieved it, and they did for many decades, it is the consequence of incredible hard work in search of equally incredible reward — monetary and metaphysical.  

The problem with defense contracting is not the popular narrative that contractors make too much money, it is actually that they make too little money. The sums are large but the current system only incentivizes spending more money to drive up the fixed percentage of profit the communist procurement policies deem reasonable. Fairchild’s financier thought it was reasonable Noyce earned a salary and had no equity. So Fairchild lost and Intel and a family of ex-Fairchild companies flourished. The USG should focus on price, not profit. If innovators can provide capability for less money, why does the government care what the profit margin is? Innovators will need outsized profits to motivate progress.  

Bill Perry and his Assault Breaker program was only possible because of the commercial success of chips. He needed a 100-fold improvement in performance to deliver that capability. Perry complained that his critics were luddites, favoring DARPA’s continued spend on its own advanced chips... but it was the commercial chips that delivered the capability.

And in the modern era, for today’s great power competition, commercial technologies will be what delivers the future of warfare. We need look no further than the battlefield in Ukraine to see that.  

Technology is a wonderfully levered thing, but there is a dangerous temptation to believe that you can change the world without engaging with its institutions, in all their sprawling entropy (setting aside Silicon Valley's frequent delusions about what qualifies as changing the world). When considering how and if to engage with national and global institutions, I've found that people gravitate toward one of two philosophies: 

1. The first holds that the world's institutions are fundamentally evil, and subconsciously people ascribe a certain malicious competence, e.g. the repressive government or ruthless corporate polluter. If this is the norm, your only choices are revolution/anarchy, or much likelier, doing nothing. I disagree, but people are free to believe this.
2. The second viewpoint holds that the world's institutions are not fundamentally evil, but need help to become great - help that must be renewed as institutions change and people cycle through. Trying to be fair-minded, people often attribute institutional problems to incompetence as opposed to malice (though in my experience this is usually inaccurate or exaggerated). 

 What I've learned from my struggles in the second camp is that there are no easy answers - but the goal is better, not perfect. These are the convictions that have helped me to contend with all the complexity that follows.

• My first principle is to get off the sidelines. Complaining is a dead end. The question needs to be “who will help them?” The answer may be you.
• If so, the second principle is to engage with the world as it is: messy and gray. 
• Overnight change is a myth. The goal is to make today better than yesterday and tomorrow better than today.

In order to bring about change, you need to cultivate some intellectual humility. But early on, it helps to not have excess amounts. Happily, most people lack humility in youth (not a judgment of millennials; it's just human nature). And this allows you to attempt epic things. Along the way, with any luck, some wisdom seeps in. As a student, I thought everything was dumb. Surely I knew a better way. As I got out into the world, I slowly realized I'd been asking the wrong question. I dwelled on the first order: why is this so broken? Eventually, I realized there was a more meaningful question: what must be true for this to make sense?  From there, you keep pulling the string until you find the underlying condition that requires fixing. 

Politics is a dirty word for idealists and engineers alike (doubly true if you happen to be both!). But you can't address the most pressing problems without also grappling with the common good, which necessitates political considerations. The reality is that politics and engineering each have roles to play:

• Politics is about accepting the tradeoffs. We live on an efficient frontier (or so we hope), forcing us to examine the tradeoffs between X and Y. In the political sphere, preference for X or Y, roughly speaking, defines left and right.
• Engineering is about innovations that push out the efficient frontier – we can have more of X and Y, and the political viewpoint merely serves up false tradeoffs that rip us apart.
• One problem with politics is that in a democracy, both sides are always right, or at least deadlocked. X is important. Y is important, too. Structurally the only good resolution is to invest in more of both. At its best, engineering can render a painful tradeoff false - think power vs. efficiency, human vs. computational acumen, or the former truism that you can only pick two among better/faster/cheaper.
• Beware thinking that engineering is a cure-all for political ills. In reality, engineers cannot recuse themselves from the painful organizational aspects of a problem they're trying to solve. Software is fairly unique as something you can open source for the public interest, but it's exceedingly rare that you get to be Johnny Appleseed, sprinkling free technology on grateful soil. The hard work happens at the intersection of products, problems, and people.

It's also worth unpacking some assumptions about the nature of institutions more generally. We tend to think of them as static, almost by definition. This is understandable, especially in the political realm. If you're graduating college this year, you have probably never known a time when bipartisanship and compromise were considered positives, not accusations to be hurled by primary opponents. In reality, though, institutions require people, and the people themselves are only as static as their motivations permit them to be.

On the flipside, one of the most important questions you can ask is whether you are working for a specific person/administration, or for the historic, enduring ideals of that institution.  This is a core principle of the US military, and regardless of one's feelings about war, it's a tradition worth emulating. In a democratic society, the failure mode is less likely a cult of personality, and more likely an institution that becomes insidiously focused on consolidating power and resources. Sometimes, institutions bloat past the point of no return, and must be broken up or declawed. Often, though, an infusion of new blood can restore focus to the real mission.

While it's clear that governments should advance causes beyond their own prosperity, I would argue this is not only possible but essential in the commercial realm. We don't remember Steve Jobs for making shareholders a boatload of money; we remember him for giving us new ways to communicate, learn, and experience art. There are many forms of value to be created, and striving for the perfect should not preclude exploring different forms of good. But forms aside, an institution's ability to do more than just enrich itself is also a great indicator of its prospects for enduring - and its worthiness of your involvement. 

On a related note, people often discount the ability to work through an institution, not just for it. I don't mean capitalizing on existing channels or infrastructure, but advancing a larger goal. You can think in terms of ceilings vs. floors. Tesla must meet certain safety and efficiency standards to qualify for tax rebates and remain street legal, but these requirements are the floor, not the ceiling. They're catalysts for more aggressive innovations, even those that don't yet have a market.

There is also the truism that if your work is black or white, it probably isn't that important - or worse, you are lying to yourself. The most pervasive problems are subtle, complex, and require deep engagement with both established institutions and emerging forces. It's easy to point to noble aims consumed by unintended consequences and be scared of getting involved. The Arab Spring fueled the rise of theocrats, not democracy. Even seemingly non-controversial developments can force hard choices. Let's say you develop a promising cancer treatment in a startup or research lab. Do you try to take it to market yourself, or through a huge pharma company? Do you optimize on keeping it affordable, or raise prices in order to enable you to expand your reach? Easy answers remain elusive, but to me, these examples are all arguments for more thoughtful engagement, not abdication. You can get off the sideline without getting on the soapbox.

Lastly, when pondering which institutions to join or serve, instead of asking “Why Institution X”, I would first ask, “Why me?” What gifts do I have to offer the world, and what kind of platform would allow me to answer the first question in the most meaningful way possible?

One of the most enduring values of Western civilization, lying at the heart of the American dream, is the concept of ownership. But in the context of material things, the virtues of ownership too often go unexamined, and the same holds true for ownership in the context of your work.

As an attempt at a definition, the only time you will ever truly own something is as an individual contributor. No one will edit the code out from under you. But as soon as you enter any sort of organizational context, you start to realize that ownership is largely a myth. The more responsibility you have, paradoxically, the more that responsibility is shared with others. From every direction, people will be editing your supposed creations.

That's not a bug - it's just the nature of the thing. As a practical matter, you need to be comfortable with people submitting pull requests against your ideas, and in fact this is part and parcel of growth. Increased responsibility is attended by both a need to drive more, and also to thrive in the shared ownership. The ability to handle this reality is a massive determinant of leadership potential, organizational health, and success as an enterprise. The inability to handle it manifests as stunted growth and pathological distrust.

Even if complete, individual ownership is largely a myth, it's still an exceptionally useful one in terms of understanding organizations and leadership. The traditional model is that you have a defined position with a defined portfolio. If this sounds stifling, it is. Its chief virtues are predictability and legibility- all other things being equal. And this is why conventional, status-driven organizations are so often ruthless and dysfunctional, more than the existence of the hierarchy itself: things change. You can only be really aggressive and formal about promoting leaders if you are equally aggressive about firing leaders. It simply isn't possible that you are always right. This is how most organizations avoid the alluring, yet messy, challenge of decoupling position and portfolio. They become factories, and as they grow, they become fractals of the old org chart.

The alternative model is the artist colony, in which position and portfolio are not only decoupled, but the position itself is an arbitrary construct. There is only the artist and the work. We see this influence in Facebook's engineering teams, where everyone from new grads to eminent veterans gets the title of “Software Engineer”. We've done the same at Palantir - with the added wrinkle that BD is all engineers as well! As a variation on the theme, recall that the Dalai Lama famously describes himself as just a simple monk.

Within the artist colony, there are leaders - and in fact, they are the rule and not the exception. Their portfolios can and will change radically over time. This decoupling allows the colony to seize more aggressively on the innumerable opportunities for leaders to drive things, without creating a corrupting permanence to whatever is being driven. In fact, I would argue that leadership can only be taken in times of flux. Otherwise, it's just more skillful maintenance. The architect of the palace does not mow its lawns, however necessary a task that may be.

The fundamental fluidity of portfolio and position constitutes the artist colony's greatest inherent advantage (and challenge). In a traditional organization, it's expected that taking away a portfolio implies the loss of position, and vice versa. In the artist colony, it's understood that this would end poorly for all, not just the artist in question. Broadly, the result would be creeping risk aversion and status-seeking. And, more often than not in my experience, just around the corner lies the perfect thing for the artist to create - while also meeting an existential business need.

Meeting these needs is not just a happy side effect of individual fulfillment, but a strategic recognition of the nature of progress. It's a constant struggle and dialogue. We need to force ourselves to the front when the things we are driving push us back, but true leaders are not content to blindly take the next hill. Each success creates the imperative to ask “now what?”. At a certain stage, the logical next step might well be an oversight role. But what's much more interesting, to me, is how often I've seen leaders ease their developmental angst and do great things for the company by pivoting from oversight to owning and executing something meaningful.

On a similar note, we're all familiar with stories of leaders not scaling because they couldn't delegate. But we should be more worried about leaders losing their ownership/execution muscle entirely. In many of these cautionary tales, I'd wager that the real takeaway should have been that Leader X didn't own/drive the right things, not that he/she didn't delegate effectively.

Another benefit of decoupling position and portfolio derives from still another paradox: despite the necessity of joint ownership (or perhaps because of it), some things simply require a dictator to get done. This is especially true in product development, where not only does halving the team often double the pace of progress, but the fulfillment of a coherent vision usually requires one actual visionary. Dictatorship as a set position would never work in an artist colony because it's antithetical to the nature of art and artists. But dictatorship can not only work, but thrive, in the context of portfolio.

It should be acknowledged that the artist colony, like the artist, is a fundamentally restless entity, and this approach is not a panacea of any kind. It requires constant engagement and examination, and betrayals of artistic principles within this world will sting infinitely worse, because so much more is at stake than status and money. Owning your work, then, is the opposite of owning your home: it's only true ownership for as long as you're actively paying it off.

Given my journey, you can imagine my first reaction to questions of work-life balance is fairly unsympathetic. I want to protest that, by legitimizing such a false dichotomy, you’re pre-empting a much more meaningful conversation. But I suspect that conversation is closer to the heart of this anxiety than most people realize. 

If you’re worrying about work-life balance at the beginning of your career, and you’re reading this, I’m guessing you’re not lazy. You’re not looking for an easy life (even if this seems like an appealing concept right after midterms). I’m willing to bet that what you’re really worried about is someone else owning your most precious possession: your future. 

Staring into the abyss of companies that glorify triple-digit hours (never mind the substance of the work), this makes intuitive sense. But having surveyed the landscape of high-tech hiring, I’m convinced you should be just as concerned about jobs that promise high stimulation and total comfort. When you let yourself be sold on easy hours, outrageous perks, and glib assurances about the project you’ll join and the technologies you’ll get to play with, you’ve just agreed to let your future become someone else’s.

I hate the construct of work-life balance for the same reason I love engineering: the reality is dynamic and generative, not zero-sum. It’s about transcending the constraints of simplistic calculations. Creating the life and the work you want are by no means easy challenges, but they are absolutely attainable. What’s not realistic is thinking you can own your future and be comfortable at the same time. Grit, not virtuosity, will be the biggest determinant of your success, for reasons I’ll explore in a bit. 

At the same time, grit and discipline aren’t enough. You need purpose. And I can state categorically that the purpose you discover, with all the sacrifice that entails, will be more motivating and meaningful than the one handed to you in the form of some glamorous project that, realistically, will succeed or fail regardless of your involvement. 

The catch, of course, is that true purpose doesn’t sit around waiting to be discovered. It requires constant pursuit. Here’s what I’ve learned from a decade and a half of sprinting.

There’s no time like now. As learning animals, we’re subject to various ages of cognitive potency. As a young child, your aptitude for acquiring a language or learning an instrument is at its peak. Accordingly, as a professional, your early 20s are the most formative stage. It is absolutely critical to make the most of this time because the pace of learning grows slower and more incremental as you age, whether we care to admit it or not. Of course, you can always learn new things, but most often the wisdom of experience is largely the result of earlier realizations having the time to compound into something richer.

The place of maximal learning is often at the point of significant pain. It’s not just about having a more pliable mind - grit, and its close cousin, resilience, are essential for taking your intelligence further than it can get on its own. And while intelligence compounds, grit degrades in the vast majority of cases. Regardless, grit isn’t something you can suddenly develop after a life of leisure. For these reasons, owning your future means choosing grit over the allure of a predictable pace.

Of course, you still need to hold a pace. Studies show that marathoners/endurance runners do tons of self-talk to push past the pain. “It’s a marathon, not a sprint” is a well-worn cliché, but it’s striking how often it’s invoked to rationalize comfort as opposed to promoting sustained excellence. Don’t think for a second that elite marathoners have trained to the point that a sub-six-minute mile pace is comfortable. It’s incredibly painful. What separates the truly elite is having found a purpose that makes the sacrifice acceptable.

At the same time, complete self-motivation is incredibly rare. It’s probably not a realistic goal, and that’s fine. Find the people who will sharpen your resolve as well as your ideas. Again, your first step matters. If you choose a job for work-life balance, chances are, so did everyone who came before. Talent is one thing when evaluating your future teammates, but ask yourself this: when you need models and inspiration to be more than you are, will you be able to find them?  Where will your gamma radiation come from?

You can find your zen in stressful, chaotic times. In fact, I’d argue this is the norm, even the ideal, for 20-somethings. Some adrenaline is good for your performance. Not having time to waste requires you to focus on the essentials and develop an innate sense of direction. That way, when you do eventually get to let your mind wander, it will be in rewarding directions. These days, I build in calendar blocks for “brain space”. That wouldn’t have made sense 10 or even 5 years ago – not because I have more free time now, but because, early in your career, you learn much more by doing than reflecting. And this can be the difference between creating your future and receiving it in a fancy envelope. 

At the limit, you probably should care about work-life balance – it’s not going to remain a static thing your whole life. But at the margin, as a new grad, you should focus on the most important problem. Find the thing that motivates you, work your ass off, learn as much as you can, and trust that today’s gains will compound well into the future – your future.

Working your ass off isn’t bleak – it’s quite the opposite. Provided there’s a purpose, sprinting at an unsustainable pace is an act of tremendous optimism. A mindset of premature retirement might sound rosy, but in truth it’s deeply cynical and extraordinarily insidious – much more so than being overpaid or overpraised, and much harder to correct.

But back to the concept of caring about work-life balance at the limit, how do you know where the limit is? Isn’t life fundamentally uncertain? Here’s what I’ve come to realize: you can’t pre-emptively retire without doing the work that makes you appreciate the chance to rest. Maybe you can, but assuming you have something to contribute, it’s going to be an empty reward. Sacrificing your potential to comfort isn’t a hedge against an early death – it IS an early death. As Emerson wrote in Self-Reliance, "Life only avails, not the having lived. Power ceases in the instant of repose; it resides in the moment of transition from a past to a new state, in the shooting of the gulf, in the darting to an aim.” 

We’ve been told over and over to choose life over work in order to achieve balance. I’m urging you, especially at the dawn of your career, to  instead choose life over balance, and make the work so meaningful that you wouldn’t want it to exist as a distinct concept. This is how you ensure that your future remains yours.

If I've learned one thing from observing great individuals (and great companies), it's that greatness is inherently asymmetric. If that sounds dangerous, it is. Any scholar of counterterrorism or cyber war will tell you that asymmetric threats require asymmetric countermeasures, but more fundamentally, they require asymmetric people. When forming a team, I don't want to assemble a polite roster of cross-functional professionals. I want the X-Men: a medley of mutants united for good.

The Incredible Hulk, in particular, embodies the growth model I've come to believe is necessary for achieving greatness. For those of you who were popular in junior high school, the Hulk began as the mild-mannered, though brilliant physicist Bruce Banner, and was transformed into the Hulk after exposure to gamma radiations from a nuclear explosion. From then on, Bruce Banner would morph into the Hulk during times of extreme stress or exigency. While the Hulk’s ability to retain Banner’s intelligence evolved over the series, it’s safe to say he was was never the same again.

So what does growth for greatness look like? It begins with accepting unevenness, and reaches its potential through a conscious nurturing of extremes. But introspection and diligence are not enough.  Real growth is scary, hard, periodic, and responsive to your environment. The gamma ray might seem like an extreme metaphor for catalyzing growth, but if you want to truly achieve greatness, it’s much closer to the reality than the safe, comfortable models we're taught to accept. You need periodic radiation, not lifting a little more weight every day. In the short term, linear development predictably leads to linear results, and in the long term, factoring in drag and the insidious effects of growing comfortable, the result is decline, as Stephen Cohen eloquently described in his conversation with Peter Thiel and Max Levchin. Intelligence is compounding all the time, and correspondingly, so are complacency and missed opportunity.

In practice, it's usually not so straightforward to go looking for gamma rays out of the gate, but there are some obvious pitfalls you can avoid along the way. One of the most important: don’t fall prey to the illusion of growth promoted by the corporate ladder.  It’s a crutch as much as a way up (and tech roles/companies are NOT immune - if you see Software Engineer I, Software Engineer II, etc, that’s a ladder). The ladder can be partially explained by convenience, or convention, but ultimately it’s there to assuage your fears – not only of not reaching your potential, but of incubating a potential that doesn’t fit the bounds. While on the ladder, you can only fall so low or climb so high. It's a false frame, not only because hierarchy is such a poor proxy for impact, but especially for lulling you into thinking achievement falls within a standard distribution.

It would be disingenuous not to acknowledge that becoming a mutant is not all upside. Make no mistake, gamma radiation can hurt you. There is always the risk of failure, and win or lose, there will be scar tissue. In that sense the ladder is also a safety net. As an aspiring mutant, you shouldn’t let false bravado obscure this realization – just recognize that in choosing the ladder you’re explicitly shorting your potential and putting protecting your ego ahead of your outcome. As an aspiring Professor X, accept that there will be failures, and that you’ll need to make highly imperfect tradeoffs on false positives vs. false negatives when hiring and developing talent.

Mentorship is likewise critical when directing mutant powers towards the greatest possible good. The X-Men would not have become X-Men without Professor X’s School for Gifted Youngsters. But again, the standard model doesn't apply. To begin with, you need mutants to mentor mutants, and in many cases, to provide the initial dose of radiation. Otherwise, even the best institution of higher learning will predictably devolve into a lemming academy. 

Once mutation is in process, one of the greatest aspects of mentorship is, paradoxically, autonomy. This is especially important because extreme growth doesn't happen on schedule, but is subject to periods of intense activity. As a mentor of mutants, you need to be attuned to these periods, and when they come, confer even more autonomy. Above all, fight your instinct to handhold (hard to do when both hands are always clenched in a fist anyway!).

The final part of the equation is to seek out the greatest challenges you can, both in terms of meaning and difficulty. And this is perhaps the greatest beauty of the gamma radiation metaphor. It's not just about unimaginable intensity. It's about an external reality leaving an indelible imprint on your internal reality. There are some gifts that are only fully formed through creative destruction, and it’s these gifts, in turn, that allow you to create new external realities - in other words, to change the world.

This post is about the current insanity in Silicon Valley, but I don't mean the valuations - at least not the ones everyone is talking about. Instead, I want to talk about how you value something much more important than common stock: yourself. 

Over the course of thousands of overwhelmingly positive interactions with top CS students over the past few years, what's scared me the most is the tendency to think of your future job primarily as a vehicle for certain types of projects. This is, in fact, one of the worst possible reasons to take any job.

In many ways this line of thinking isn't so surprising. Perhaps because the long-theorized tech crash hasn't happened, and most companies (even relatively innovative ones) think of hiring as filling slots, our economy continues to promote skills over aptitude and ability. And even the best schools are much more effective at teaching subjects than synthesis. As a result, even in an age when software engineers are starting to be properly valued, there is a real risk of being commoditized - ironically, by yourself. 

Apart from an earnest desire to cultivate "valuable" skills, however, is something I'll call techno-hedonism. Besides just thinking of your job in terms of projects, this means evaluating projects by how pleasurable they are to you versus how much good you're creating in the world. As a result, topics that could be invaluable as part of a greater whole - especially things like machine learning - become playthings. And this is how young people who honestly thought they were going to change the world end up being paid too much to quit to serve ads more effectively.  In the degenerate case your employer becomes something to be agnostic about, merely a vehicle to work on a specific project of hedonistic desire.

Rather than deciding based principally on the project, I would suggest there are two questions that should inform everything else: Do you believe in the institution? And do you believe in yourself?

Evaluating the institution involves many more questions, but I'd argue these few are most important: Is there a real opportunity to make a positive impact? If so, is the team equal to the challenge, or (more likely) on the path to getting there? Is there a possibility of surviving as a standalone entity - this is almost impossible to know ex ante, but if the stated goal is to get acquired that should tell you something. Do they have a real mission and culture, or just hedonism and homogeneity? Do they invest in an individual's growth, or just increased productivity? 

By believing in yourself I don't mean projecting an arbitrary level of confidence - it requires a willingness to critically assess your strengths and weaknesses and reconcile them with an emerging and constantly evolving sense of purpose. This cannot happen overnight. If you're betting on your ability to do something important, you'll learn - piece by piece -  to intuitively subordinate the process to the goal, and separate the act of discovery from the procedural. By contrast, if you're betting on your ability to stay fulfilled by repeatedly doing a series of tasks, however pleasurable, you're actually shorting yourself. 

It's not so difficult to see the surface characteristics of an institution for what they are - when you become enamored of a slick office space, at least you know you're being shallow. Becoming enamored of projects, on the other hand, feels like investing in your most important assets when in fact you may be stunting them.

I want to emphasize that this is not happy talk. It is unbelievably hard work. Having it all figured out now is the unrealistic part - and if you actually do succeed in your design, that's when the reality often proves to be bleakest.

Engineering is fundamentally generative. Specific implementations may be highly deterministic, but the defining character of the work is possibility. It's understandable to want to cling to certainties, especially after hearing what a dark and chaotic world it is for most of your conscious life. I say: embrace conscious ambiguity. The alternative is a predetermined future - one that truly belongs to the robots. You are not a lottery ticket - but neither are you an algorithm.

 

“Do important things” is often invoked as a rallying cry in these pages, but this time I want to talk about something more important than innovation, invention, entrepreneurship, and all the rest. I want to talk about dharma. More specifically, I want to talk about your dharma.

Classically speaking, dharma represents both cosmic law and order – our universal duty - as well as reality itself. Upholding your dharma, then, refers to both your ultimate responsibility, and upholding the truth.  It is no accident that I say your dharma. The truth, while in one sense absolute, is also deeply personal, and rooted in the enduring power of the individual.

With commitment to the truth as the first principle, your code of conduct is simple: When you see something that's broken or bad, you have to say something about it or fix it yourself. Just as importantly, when you hear something, listen. It’s not just about the success of the organization, but also a moral imperative not to let anyone you care about fly off a cliff.

In practice, this is extremely painful. Honest, unadulterated feedback is as emotionally alien as it is intellectually obvious, whether giving or receiving. Confronting the truth together is a social endeavor, yet it flies in the face of all social convention and pleasantries. Unlike you or me, the truth doesn’t have feelings – but that is precisely why it’s the truth.

Of course, it’s easier to face hard truths when we talk about collective failures. These are important to address, and can be invaluable object lessons for the organization writ large. Individual failures, however, are the ones you, and only you, can control. Accordingly, the most painful and most vital incarnation of the truth is individual feedback – all in the service of discovering and fulfilling your dharma.

This matters on multiple levels. In practical terms, nothing happens unless you make it happen. Day to day, the bias towards action is one of the most valuable things you can institute.  Without your concerted action, things like planning, analysis, strategy, et cetera are just distractions from an empty center.

However, dharma is also about the unlocking the essence of the individual. Facing your dharma means stripping away the pretense, delusion, and distractions to reveal who you are and what you are meant to be doing. You uphold your dharma in the service of both the individual and the collective. For the whole to be greater than the sum of its parts, the parts cannot seek anonymity and cover in the whole.

Likewise, true feedback comes from a foundation of investment in the individual. The underlying intentions need to include the opportunity to grow from mistakes and the willingness to help someone get there. We all like to talk about investing in people, but it’s important to internalize that hiring isn’t the end of the road. The hard part starts after - especially for the most innately talented individuals. If you don’t give them feedback, you’re just as guilty of coasting on their talent as they are, and you will inevitably reap the consequences.

As many a wise master has observed, there are countless paths to dharma – indeed, there are as many forms of dharma as there are seekers. Everyone arrives at the truth in a different way, as evidenced by leaders as diverse as Ray Dalio, Prof. Carole Robin, and Peter Thiel.

Ray Dalio’s Principles is more than required reading at Bridgewater, and Bridgewater’s culture of “radical transparency” is almost infamous for the degree to which honest feedback is emphasized. Dalio’s most basic principles states: 

“Truth - more precisely, an accurate understanding of reality- is the essential foundation for producing good outcomes.” 

It seems simple enough, but the real genius of Principles is how he mediates between the truth as an absolute and the individual experience: 

“Above all else, I want you to think for yourself - to decide 1) what you want, 2) what is true and 3) what to do about it.” 

Dalio also caveats that “you can probably get what you want out of life if you can suspend your ego”, and the same can be said of feedback. For most of us, this will be the hardest battle.

One of the more pernicious, and also subtler, difficulties of governance is something I’ll call the tyranny of optics. Across the organizational spectrum, you find systems that are designed to appear transparent, fair, and free of conflicts of interest. Yet all too often, the result is gridlock and bad outcomes for the honest actors, while actual corruption is only pushed deeper underground. It’s the ultimate bitter irony: instead of functional compromise, you get institutionalized disaster.

The legacy government acquisitions system is a perfect example. The driving force is typically not a desired outcome, but rather a long list of requirements established to pass the eye test. The unintended consequences of these requirements, combined with their tendency to stifle innovation, result in the worst of all possible worlds - for the mission, the taxpayer, and the many people doing their best to both produce and acquire high-quality technology.

One of the greatest pitfalls is contracting on a cost-plus basis. This is largely a function of optics, as well as the inherent difficulty of placing value on high-tech innovation (and the age-old confusion of cost with value). The problem is that a fixed profit margin means you can only make money by increasing revenue – there’s no incentive to increase efficiency, even though efficiency is the whole basis of Moore’s Law. In essence, you substitute accounting for accountability, and the effect is that the true value of technology, and the true potential for innovation, are obscured by the very mechanism meant to ensure transparency. It’s also worth emphasizing that for the vendor, it’s about simple math, not corruption. When you can only make money on the top line, a rational actor has no choice but to conform or find a different business.

Furthermore, the system is designed to evaluate the surface qualifications of a vendor to perform work at the government’s risk – have they done something like this before for similar clientele? When building massive hardware investments such as aircraft, this might seem like a reasonable question (though the success of SpaceX has chipped away significantly at the conventional wisdom). When applied to information technology, it’s much more obvious what an arbitrary standard this is - imagine if Larry Page and Sergey Brin had been subjected to these considerations when they were raising capital. The consequence is that the number of “qualified” contenders remains flat over time. This, in turn, creates in an anti-competitive vicious cycle where the presumed ability to deliver is based on perceived qualifications, rather than those qualifications being based on the actual ability to deliver.

Of course, technology projects fail all the time – but because optics are paramount, there’s no willingness for the customer or vendor to admit failure. Instead, we keep sinking money into the same projects until any resolution seems palatable, or the original need is forgotten. Paradoxically, the system demands perfection, yet actual failure is shockingly acceptable – so long as the vendors are “qualified”. Because these failures are overseen by familiar faces, the vetting committee still boasts a perfect record. It’s like a dystopian version of Blackstone’s formulation: better ten credentialed companies should fail than one startup. Consequently, no one is willing to take the kind of development risks that could yield transformative discoveries. Failures that amount to sunk costs are acceptable, while the ones that could really teach us something are unthinkable.

A highly respected veteran of Congress and the Executive Branch once told me that one of the more underreported challenges of DC was that killing earmarks only removed much-needed grease from the system, predictably causing the machinery to grind to a halt. Ironically, earmarks connoted a certain honesty because everyone knew what was going on -The practice allowed for plenty of valuable give-and-take - the real problem was that in many cases the optics were just too shaky.

Since the earmark moratorium, we’ve been treated to an endless game of budgetary chicken that has certainly led to worse outcomes for taxpayers than earmarks ever did. Meanwhile, conflicts of interest haven’t gone anywhere – they’ve just reappeared in the form of more insidious slush funds and legislative blackmail techniques. Technology acquisitions and Congressional deal-making might appear to be very different beasts, but in both cases, the substance of compromise and pragmatism has been replaced by the rigid ideology of covering your backside at all costs. When optics are the primary concern, you can’t even have token cooperation, let alone the partnership needed to solve hard problems.

Bill and Melinda Gates’ recent Wall Street Journal editorial, Three Myths on the World’s Poor, exposes the tragic result of focusing on optics above everything else. Only a small percentage of foreign aid is lost to corruption, but that part always receives vastly disproportionate attention. If the absence of any perceived impropriety became the design criteria for providing aid or philanthropy, we’d only hurt the very people who need the most help. As the authors poignantly ask, “Suppose small-scale corruption amounts to a 2% tax on the cost of saving a life. We should try to cut that. But if we can't, should we stop trying to save those lives?”

The tax metaphor also helps to expose the rampant cynicism that preys on optical controversies. Almost no one would consider a small tax, or other nominal costs of doing business, a good reason to abandon an overwhelmingly profitable enterprise. Why should the criteria be impossibly strict when we stand to gain lives as opposed to dollars? Perhaps better than anything else, the humanitarian aid challenge reveals the logical conclusion of elevating optics above everything else: since a perfect solution is impossible, we’re better off doing nothing.

Every election cycle, someone promises to run the government like a business. Setting aside whether this is desirable or feasible, the obvious challenge is that the optics become most restrictive when the government bears the risk (as businesses generally do). Yet vast opportunities exist for government to transfer risk from taxpayers to suppliers. Imagine a marketplace where vendors can only compete if they guarantee an outcome or your money back. Optics would revert to their proper place: still a factor, but far from being the first or only consideration.

By ending the charade of demanding perfection, we can stop wasting time on the fantasy of eliminating risk and instead focus on the real work of managing it. When you practice the art of the possible, paint will inevitably splatter – but to a realist, the result is infinitely more attractive than an ideal that will never be achieved.

Without commenting at all on the policy wisdom of the Affordable Care Act, it’s clear that the rollout of Healthcare.gov has been disastrous. This has been chronicled more diligently elsewhere, but can be summed up by noting that, while Healthcare.gov was plagued with bugs, crashes, and general confusion, a team of three college students replicated much of the desired functionality of the site in a few days.  Of course, the alternative site, HealthSherpa, does not handle the user or data scale of healthcare.com or perform the most complex operations of applying for coverage, but the contrast between a site built for free and the ~$600+ million obligated for healthcare.gov is sobering.

We can draw a few lessons from this affair.  The first is that it represents a deep structural problem of government IT projects. The process used to bid out and build healthcare.gov was not, contrary to what you might have heard, especially unique or nefarious.  On the contrary, it represents the norm for large federal IT projects: mandating what should be straightforward products to be built from scratch in many ponderous phases, replete with massive sets of requirements and a commensurately high number of billable hours. 

The major difference is that this time, the users are the American people. The frustration of grappling with subpar technology is the same experienced daily by some of the most vital people in our public service ranks.  Soldiers, intelligence analysts, law enforcement officers, and veterans care workers, to name just a few, are routinely forced to implement tools that are barely functional, told to simply “make it work”.  This is by no means meant to minimize the headaches associated with healthcare.gov – on the contrary, it points to the need for real, systemic change.

There are two fundamental flaws at work in the legacy government IT acquisitions model. The first is that the same procedures used to acquire tanks and aircraft carriers are used to build software. Yet software development is by nature a creative, generative, iterative process, not a static set of requirements that won’t change significantly over the lifecycle of the product.  And while good software is never truly finished, the essential building blocks can often be delivered right away - the key is that you’re creating a basis for iteration and creative enhancement, not obediently following the same blueprint for years at a time.

The second, and subtler, flaw is the failure to recognize that America in general, and Silicon Valley in particular, are unique in the ability to build software.  Many remarkable advantages of American life have contributed, in turn, to our dominance in software development. Pondering an increasingly data-driven future, our abundance of software talent has to be considered one of America’s most strategic resources, and leveraged and fortified accordingly. Sadly, in the current IT acquisition landscape, armies of contactors are paid by the hour to produce a crude facsimile of what our best software artists could create for a tiny fraction of the cost - but ignoring such a precious asset would be a mistake at any price.

One great irony of the healthcare.gov fiasco is that a major rationale for the Affordable Care Act was the idea that Americans can do better than the legacy healthcare system – only to see what should have been a slam-dunk website rollout crippled from the beginning by the IT acquisitions machine, another legacy system. Regardless of one’s views about the law itself, though, one saving grace is made clear: if we want to do better, doing what we’re already the very best at seems like a good place to start.

One of the most fundamental human desires to believe that something is either A or B, and many complex endeavors are compromised from the beginning by treating the A/B split as a first principle. Binary logic may explain well-understood processes, but eventually the old rules cease to apply, as with the failure of classical physics to explain phenomena at atomic and subatomic scales.  To understand quantum theory, you have to accept the wave-particle duality, and even then, it turns out that no one really knows why light exhibits both wave and particle properties.  We can observe, even predict, but not quite explain.

Startups are subject to similarly misunderstood dualities.  Simple minds want to know if winning depends more on doing A or B:  Should we move fast, or ship quality? Build footprint or monetize?  Optimize on breadth or depth?  The winner, however, realizes that you have to figure out a way to do both.  How this is accomplished is highly contextualized in practice, but it begins with the realization that you cannot have one without the other and hope to succeed.  If it were as simple as doing only one thing well, the success rate of venture capital would be much greater than 10%. And when you do succeed, as in quantum mechanics, recognizing that things work a certain way is more important than knowing why (for the purposes at hand, at least).

A venture also displays both continuous and discrete elements.  From a wide angle, the growth curve or product lifecycle may resemble a wave function, but it’s also extremely iterative, and is most efficient when individual iterations occur at consistent intervals.  Likewise, one characteristic is often expressed through the other, much as particle emissions are dependent on wave functions. The focus and abstraction needed to go broader also allows you to go deeper effectively.  Similarly, in the course of developing a vertical solution, you often end up sharpening your intuition about how slice the problem horizontally.

When striving to achieve both A and B, you often need to consciously set up opposing forces to achieve your goals.  For example, you need hackers who are relentlessly focused on solving the customer’s problems, even if they’re comparatively poor at productization and long-term code stability, and you need artists who are relentlessly focused on productization and pristine architecture even if their sense of customer urgency leaves a lot to be desired.  How you make them work together productively is an art - there is always some violence, but it starts by recognizing you need both, and accepting that their interactions only need to be productive, not harmonious.  The results of this type of particle collision are very difficult to know ex ante, so the safest bet is to find the best exemplars you can of each type – people you would want to work with individually.

The need to harness opposing forces sometimes extends beyond types of goal orientation to personality types (though these often go hand in hand).  Again, it’s up for debate why this is the case, but the anecdotal evidence is extensive.  The classic example from quantum physics is Richard Feynman and Murray Gell-Mann’s collaboration on the theory of beta decay.  Feynman was famously mischievous and irrepressible, while Gell-Mann was almost painfully serious and methodical.  While they frequently found each other exasperating, their tension was tempered by strong mutual respect – an obvious but sometimes overlooked component in organizational design.

Conventional high-tech wisdom posits that among the qualities of “better”, “faster”, and “cheaper” you can only pick two.  With the right team, you can do extraordinary and counterintuitive things. You can be better, faster, and cheaper – you just can’t be better, faster, cheaper, and also comfortable, which is the true contradiction. At the risk of resorting to truisms, doing hard things is hard - comfort is simply not part of the equation.  As Feynman himself once quipped, “You don’t like it, go somewhere else!”