The Right Kind of Crazy: A True Story of Teamwork, Leadership, and High-Stakes Innovation

by Adam Steltzner

I resolved that the solution for my anxiety while staring at the unanswered question was to remain calm in the presence of the openness, to not close off the inquiry too soon and thus run at full speed into a solution that might not take the whole truth of the problem into account. Holding on to the doubt meant listening to all that the problem had to say and not making assumptions, and committing to a plan of action based on them, until the deepest truth presented itself.

As I look back on that frustration, I can see a rich cocktail of youthful naiveté, headstrong certainty, and a dash of genuinely useful observation about the way the organization operated. I’ve since learned that this mix is fairly common in intelligent young employees first stepping into a job for a big firm, be it JPL or another aerospace firm, Apple or Google, or anywhere else. A fresh perspective frequently brings with it insight not available to those on the inside. Zen Buddhists call it shoshin—“beginner’s mind.” That is one of the greatest gifts an entering employee can give, but it also presents a challenge. The same vision that sees the holes in the operation does not see why those holes might have been accepted long ago as a necessary evil or even put there by an act of will. Beginner’s mind presents both challenges and opportunities for the institution and for the new employee.

Here’s what I wish I could tell my younger, impatient self: Advancement and influence in any industry do not in general keep pace with an industry’s most famous outliers. At least in a meritocracy, such as JPL, if you do good work and really focus on mastery and excellence, good things happen, for the institution and for you. That is not to say that “the institution” is a benevolent, all-knowing, and all-rewarding entity, but any institution desperately needs good people. There is a vacuum at the top. Their desire for good people and talent is insatiable. If you do good work that is valuable to the institution, you will inevitably be vacuumed upward.

What’s more, true authority comes not from a title or position but because your words are well thought out, or at least strive to be. Nothing puts more weight into your opinion than that it is well considered, well articulated, and coming from a high performer. If you want people to take you seriously, better put your head down and be useful to them.

True learning lies in the latter approach: to embrace the confusing mess of tangled thought and tangled algebra and let curiosity loose to hunt for what is right and true. This is not a process without anxiety. It requires that both parties sit side by side with doubt and hold on to it no matter what the ego demands.

One of my favorite Teddy Roosevelt quotes is “Far and away the best prize that life offers is the chance to work hard at work worth doing.” I feel very lucky to think that the work I do is worth doing. I also feel very lucky to work on a scale that requires people to work together in teams. I appreciate teamwork so much that I’ve developed a riff on Mr. Roosevelt’s quote: “That great work requires many people coming together is one of the great prizes life has to offer.”

Bringing all that you have requires offering up your opinion in the absence of an invitation. It requires self-authorization. You need to believe that you have the answer, and you need to give it to the team, even if you only think you have the answer. It is a form of leadership, and it is needed at every level and in every element within a healthy and high-functioning team.

My head spun when I looked at the breadth of the problem. It was hugely complex. As Chris and Randy worked to understand the penetration problem, it was not at all clear how we would ever converge on a solution. We hadn’t broken the problem up properly and divided it up among ourselves. What things did we have to test? What could we understand from basic physics and analysis? What pieces did we need to engineer? And what were the central questions we had to answer? There seemed to be a vacuum of technical leadership in the way we were attacking this.

started from the perspective that I was definitely not the boss or anything like it but, rather, that I had thought about our problem and broken it up into the pieces that we could analyze, the pieces that we needed to test, and the data we needed to get out of our testing. I shared my reasoning with them, and they agreed to follow the plan of work I proposed. The lab has a long-standing tradition of yielding to the strengths of arguments, but this was the first time I had stepped up and offered my arguments for consideration.

I wasn’t making any particular claim to leadership, nor had I been given prior authorization to lead. I’d acted like the team’s leader by asking Chris and Randy questions that would help them organize their thinking and solve their problem. That’s what Dara liked. His promoting me endorsed the idea that you have to rise to the performance level of the job you want to have, which means that you have to just freakin’ do the job. If you come in acting like the boss, bossing people around in order to feel powerful, the most common response is going to be “Who the fuck are you?” So the trick is to exhibit leadership—lead—without having to claim leadership or to subjugate others. This is leadership really as a service function, as a gift to the group.

The converse lesson is that anyone who waits around to be authorized to speak up is never going to be a leader. It’s just not in their nature.

His intonation was almost an en garde! It certainly was not polite and deferential, nibbling around the issue. It was more like punch and counterpunch. I liked it! “Oh, we’re going to use fingers,” I said, referring to the curved-leaf spring-like steel tines that would stretch out and keep contact with the surface as the comet material receded from the spike. To which he replied, “Well, are you just going to rely on the heat path along the spring length?” And so it went, punch and counterpunch, back and forth. These direct, assertive, full-on encounters made the other groups seem passive and risk-averse by comparison, as if they feared the truth and the changes it could bring. These Champollion guys, who had been the Pathfinder guys, were more like the naked barbarians I’d heard about as part of the foundation myth of JPL, brawling for the truth around the campfire.

At which he said, “No. You look. I don’t need this,” and stormed out of the room. Everybody who saw it was stunned. He was like the guy who pulls the pin on the grenade with the handle down and says, “Do what I say or else!” His storming out was like letting the grenade explode. It’s upsetting to confront that kind of emotional brinksmanship; even more so to go over the brink! To me it felt like I’d seen what too much emotion could look like. We solve problems at both the conscious and subconscious levels. Dara was and is a naturally intuitive leader and problem solver. His emotion was there to protect the space he needed in order to problem-solve in a kind of subconscious, perhaps even slightly irrational, way. In an engineering environment, this combination of the rational and the irrational is so unusual that it shakes people up. That makes it an incredibly powerful catalyst for getting at the truth, but like any powerful ingredient, it needs to be used sparingly.

Cow is data without a model, theory, or connections. Bull is a model with no data to support it.

It is not enough to log and tabulate the observables; it is essential to develop an actionable understanding of the underlying causes of the observables.

An organization needs to reflect the people in it. We are not interchangeable human units of effort or skill. Each of us is unique, and the organization needs to be shaped around us.

Not wanting to be held hostage to talent that might walk out the door, most organizations chafe at the idea that people are unique and that any one person’s contribution might be crucial. Institutions want to believe that policy and process, not individual belly buttons, determine the quality of their product. Thankfully, JPL, more than most, accepts that policy and process without the right people is a losing proposition.

As a result, the lab hadn’t staffed the project with architects—that is, the kinds of engineers who can look at a blank screen and begin to envision something radically new, or even question the plans in front of them. Line management had simply brought on engineers who were the very best at bolting shit together and testing it again and again until they were sure that every component was perfect.

As a rule engineers are inclined to dig deeper than the average person out of natural curiosity as well as practical necessity. But sometimes we hold back, usually when we fear the implications of what we might find, which might lead to things we might have to change.

Sometimes a project needs the manager who will keep it moving, and sometimes it needs the troublemaker who will stop and question.

I think most people in positions of leadership have some insecurity about their authority, at least at some level. I definitely did, and the way I cemented my authority, at least in my own eyes, was to push to understand my teammates’ jobs as well as they did. I tried to win authority through technical credibility. I was able to think of the problems and the solutions to the problems my people were working on right alongside them, which seemed to put them at ease. It also meant that the team meetings had a great deal of intellectual back-and-forth as I encouraged everyone to think about one another’s problems and solutions for them. This was great fun for me, and I think it helped the team grow closer to one another and more engaged than they had been before.

a leader doesn’t have to be liked by everyone on the team, and probably shouldn’t be. There have to be opportunities to bitch and snipe about the person Mother-Goosing your effort—a chance to vilify, wrongly or rightly. I didn’t give the team room for that.

The demands of the task drew the team into a more complete collective understanding of their mutual effort, and really knowing how your component fits into the bigger system is reassuring. It also makes your work all the more meaningful. It invigorates and inspires. Ultimately, it’s absolutely essential to getting it right.

Systems engineering is all about understanding the essential technical behavior and the underlying physics of the spacecraft’s functions and how those connections and interaction can be as important or critical as the behavior of any one element. It’s understanding the risks of each piece and the emerging behavior of the whole system, knowing what can be known about the system with certainty and what can’t, how to balance the risks, and how to make design choices to get a complex system to do what you want it to.

I started to recognize that when you are system-engineering, you are really engineering not only the engineering system but also the human system that creates it.

I started stepping back from my material (PowerPoint slides, usually) and thinking about the essence of what I was trying to say and why I was trying to say it. Then I started thinking of the counterarguments to the position I was advocating and considering whether they were more valid than the support for my point. In short, I started doing what I knew Richard was going to do as soon as I got in the room with him. That made my work better, more balanced, more deeply considered.

One of the problems with space exploration is that we never have enough iterations to allow us to fully learn from our mistakes. When it takes five years to launch and then two more to realize success or failure, there’s a time constant that defeats learning. By the time the deepest truths emerge, most of the humans who were in on the creation very likely have moved on. If we hire a contractor that saves us millions but in the end their work fails, it’s hard to know what caused the failure, because we don’t do enough missions to gather the right kind of statistics, so we’re left with too much supposition.

In some senses what we do when we develop a one-off spaceflight system every five to ten years is akin to a modern high-tech start-up. It takes a few years, every time it’s different, and you don’t get a lot of practice.

All of this is why asking the right questions in the first place, and then listening deeply to the answers, is vital to embarking in the right direction.

Well, to start, the right questions are those that are essential and profound, which means those that are architectural in their dimension. Frequently the right questions define the watershed between differing strategies of solutions or crisply define an essential risk.

In fact, it is our curiosity that we need to engage and our fear that we need to hold at bay, or release, if we’re going to find what we’re looking for.

It is as if there are two forms of decision making: fear based and curiosity based. In fear-based decision making, we find ourselves wanting the answer as fast as possible. We don’t really pay attention to, or care, what the answer is—we just want something, because the open question makes us anxious and fearful. In curiosity-based decision making, we use one of the core traits of our species to pull apart, examine, and wade into the open question. In my experience curiosity-based decision making yields much better solutions.

What I found to be most honest and most effective was to have all the groups I interacted with good-naturedly bashing one another. The goal was to find a way for ideas to win rather than people, so it was important to set up this jousting as a test of competing truths rather than competing egos or hierarchies.

To get to that place where the conflict could be creative rather than combustible, we also needed the team to remain very close personally, so I saw part of my job as being a kind of high-tech camp counselor. Growing up in Marin in the seventies, I learned from my parents early on to be open minded and accepting, but now I realized the need to take it further. I wanted to genuinely like the people I spent my days with, if for no other reason than that it makes life more enjoyable. We work with our teams for eight to twelve hours a day, five days a week—a huge chunk of our waking life. So I set out to try to find one thing to love about everyone. Sometimes you can’t get all the way to loving the whole person, but maybe you can love the goofy way they show up in a bike helmet, or the way they have those cat pictures all over their work space, or dress fastidiously, or love to organize thoughts on spreadsheets, or how they can’t help dreaming of solutions that appear off topic. Oddly enough, in time the love actually happens.

In the same way, your frustration with someone doesn’t have to be with him or her fundamentally as a person. You try to limit your frustration to just the behavior, because getting to the hard truths requires people who can show affection and respect for one another at the same time that they feel ...

The crazy part is being willing to be wrong in search of right and not worry about looking foolish.

The key to searching for the truth is to hold passionately to your beliefs while simultaneously not feeling entrenched in your position, to be able to let go of the need to defend it in order to save face. It’s almost a Buddhist thing, where you’re not necessarily free of ego and concerns about status, but you’re able to sit with them and maintain some objective separation. It’s about letting ideas win, not people. It’s about finding what’s right, not being right.

Holding on to the doubt means you listen to the problem until the deepest truth presents itself. But that, of course, raises the question: How do you know when you’ve gotten down to the deepest truth?

Then there’s your Spidey sense, the confidence in your gut that you learn to trust beyond the horizon of the hard data. It’s based on the data you have, but it also encompasses confident extrapolations. As such, Spidey sense is the essential counterbalance to the rational minimalists who deconstruct and analyze every argument into its elements and miss the intuitive altogether. It is the part of our awareness that is reaching out to get a grip on the unknown unknowns.

My adviser at Caltech was a fabulous Scotsman by the name of Ron Scott. He used to say, “Your brain is like a sausage casing. It can only hold so much, and if you try to stuff in too much, something important might get pushed out, like your daughter’s birthday.”

Once again I was peering into the chasm where the bedrock truth of the hard data—the math and the physics—did not hold the answer. They are incredibly precise, but they are merely the tools we use to build our models, so you can’t blame them when the model fails. In this case the only way forward was to exercise those particularly human capacities of engineering judgment, and maybe Spidey sense, as well as an activity called risk leveling, which is the essence of Entry, Descent, and Landing.

That path of sticking doggedly to a solution without continuing to ask if it was the right solution cost us at least two years on the sampling system alone.

I was too afraid to appear indecisive, too concerned with making measurable progress, and not willing to question.

I feel that many leaders walk a difficult line between these two poles. On the one hand, we have to question every assumption, dig into every truth. But questioning can also go too far. So when does holding on to the doubt cross over into paralysis by analysis? And when does an attempt at being intuitive and instinctive lead instead to just being half-baked and wrong?

Breaking the matter down, understanding your state of understanding, and keeping your mind free and purely focused on the matter at hand—not yourself—is the only way to strike the right balance between consideration and action. You have to understand the problem in terms of those Rumsfeldian logic boxes: the known knowns, the known unknowns, and the unknown unknowns. You have to get your arms around everything you understand about the task at hand, along with everything you know you don’t know. But you also need some sense of the magnitude of what you can’t even directly sense. You have to be personally invested, but you also have to be detached. You cannot be bound up in concerns about your own performance or what people might th...

The people who manage the early development of a project are often chosen for their ability to keep things moving without being absolutely certain they’re moving in the right direction. Early on it’s enough to simply be advancing, exposing ideas, stirring up possibilities. Later you need people who are going to be brutal on sorting through the ideas and incredibly precise on directionality.

We humans are an extremely social bunch, and most of the great things we have done have been done working together in teams. How we organize when we work together can lead to greatness or disaster, and although the science of organization is studied and well developed, in my experience organizing humans together for a common goal is much more of an art—creative, fuzzy, and emotive.

I’ve learned to think carefully, when I’m assembling a team or welcoming a new member, about the entire cast of team members, their wants and desires, stated, unstated, and sometimes unconsciously demonstrated. I consider the strengths and weaknesses of everyone, including myself, and how each might grow or improve. With all this information, I take the traditional project organization structure for the portion of the project I am involved in and I warp it, distort it, and make it conform to the specific individuals involved. The obvious goal is to create the most productive team in order to deliver the best team product.

This idea of strengths and weaknesses, dislikes and desires, holds at an organizational level as well. Groups, subteams, and even institutions take on a personality, in a remarkable, complex, and sometimes unfathomable way. IBM has a different culture than Apple, which is different still from Autodesk, or Cisco or Google. Within NASA the Johnson Space Center has a different culture from Ames Research Center, which differs from Marshall Space Flight Center and, of course, from JPL. The differences among the various centers’ cultures can be used to forge strong, highly functioning teams, or they can lead to disharmony and discordant results.

Finding how to help each of us as individuals, subteams, and/or institutions play to our strengths, grow toward our desires, and maximally contribute to the effort of the whole is the essence of leadership.