In Lane‑Breaking Part 1, I told the up‑close story of lane‑policing and psychological safety from inside teams. In When One Lens Dominates and Safety and Stewardship, I zoomed out to talk about bias, culture, and leadership. This final part is more personal and forward‑looking: how this frames DEI and cultivation. What it means to be a polymath in a specialist world, why small, sharp teams need lane‑breakers, and how critical thinking and AI change the game for the next epoch.
This is longer than I planned. I’d rather land the plane here than drag everyone through Part 5.
I know parts of this manifesto can read as pessimistic or “too negative.” That’s not an accident. I’m critical, not cynical: I’m naming failure modes I’ve seen up close, and that is never going to sound like a recruiting brochure. If a line stings, my ask is simple: treat that discomfort as data, not as a reason to look away. The only reason to bother writing any of this is because I think it’s worth addressing, and if a manifesto about culture never pinches anyone’s nerve, it’s probably not saying anything that matters.
DEI in the context of lane‑breaking
When I say I’m pro‑Diversity, I mean who’s in the room and what perspectives are actually present, across identity, discipline, background, thinking style, and lived experience. It’s the mix of people and lenses you’ve got before anyone opens their mouth.
Equity is how the bar gets applied. It means everyone has the same opportunity to meet the bar and raise it: the same access to information, mentorship, feedback, and stretch work, and the same expectations when it comes to standards. It is not lowering the bar for optics. It is making sure the path to the bar isn’t quietly blocked for some people and greased for others.
Inclusion is whose voice really counts once they’re in the room. It’s whether people can speak with candor from their lane and across lanes without being punished, dismissed, or tokenized. It’s whether discomfort from marginalized folks gets treated as real data about the system, not as “being difficult,” and whether people who aren’t in the dominant group ever get to shape the actual decisions, not just the branding.
For me, DEI and lane‑breaking are the same fight at different scales. Diversity without trust across disciplines gives you a curated cast. Equity without consistent standards gives you scapegoats. Inclusion without the freedom to speak across lanes just ensures the uncomfortable voices stay out of your way. Real equity and inclusion require cultivation, not just slogans and hiring targets.
Cultivation as a deeper kind of kindness
Cultivation is helping something—or someone—grow into what it could be. It is a hard form of kindness: slow, inconvenient, and sometimes uncomfortable for everyone involved. You do not get resilient teams, deep craft, or honest cultures by protecting them from friction. You get them by exposing them, with care, to the right kinds of pressure over time. This kindness is Punk.
This is where “nice” and “kind” quietly part ways. Nice keeps the surface smooth and the room relaxed. Kind is willing to introduce tension in service of the person, the team, or the work. Nice says, “Let’s not make this awkward.” Kind says, “This will be awkward, and it still needs to be said.”
Toxic positivity and ruinous empathy are the failure modes here: one lets the performance of optimism replace contact with reality, the other withholds the very feedback people need to grow.
If you care about a person, a team, a community, or a studio culture, you will sometimes choose the harder path now so they can be stronger, freer, or more capable later. You tell the uncomfortable truth. You hold the bar steady when it would be easier to let it sag. You give mentoring, repetition, and correction you are not in the mood to give. Cultivation does not always feel kind in the moment — it is what kindness looks like when it takes responsibility for the future, not just the mood in the room.
That might sound idealistic, but the outcomes it’s tied to are very concrete.
The research behind all of this is real, additional sources are collected at the end of this post if you want to go deeper.
If psychological safety, equity, and honest cultivation are the conditions for real work, polymaths are some of the people who make the most of those conditions, and suffer the most when they’re missing.
Polymaths in a specialist world
Part of why this hits so hard is that I’m not built as a narrow specialist — that’s the full story, and it’s in Part 1. What I’ll add here: in a world where AI lends you IQ points, compresses learning curves, and cheapens narrow implementation, the ability to see across lanes is only going to matter more, not less.
To someone who only understands specialists, that still doesn’t look like value. It looks like wandering out of your box. Lane‑policing is the system’s way of asking you to stop, to play the smaller, simpler character on your business card.
If your studio says it wants innovation and big bets but treats polymaths as threats instead of assets, you do not have a talent problem. You have a lens problem. The places where that lens is already different (the ones that treat cross-lane people as a feature, not a bug) tend to look a lot like this:
Where this way of working fits: small, sharp teams
My first game was a handful of us in a loft, figuring it out as we went. Nobody had a narrow lane. The “graphics person” might also be the tools person. We didn’t even have an “engineer” – we were hacking in HyperCard. The level designer might also be scripting interactions. The 3D artist creating a room might devise the “puzzle”. Whoever cared the most about the schedule that week was effectively production. We were not doing that because it was trendy. We were doing it because there was no other way to ship; there was no education that could tell us how to do it “right.”
Today’s AA and indie teams are not that different.
They are small groups with limited budgets and big ambitions. They win by moving faster than bureaucracy, letting ideas come from anywhere, and trusting the people who can see across systems and call out problems early. As AI picks up more of the rote work, the real leverage in those teams is the people who can glue everything together. They do not have room for fake safety, for leaders who only want good news, or for lane‑policing as a management tool. When there’s no fat in the system, you either make it safe for people to tell you the ugly truth across lanes, or you ship something the market will tell you the ugly truth about later.
The stuff that breaks big, rigid organizations – polymaths, multi‑passionate weirdos, cross‑disciplinary critique – is the stuff that makes small, sharp teams dangerous in the best way.
I’ve seen it work at scale. At Lightbox, we needed four times the worlds, each four times the size. The real solution was a content and pipeline problem: design, art, engine, and performance iterating together, each feeding back into the others. Nobody could stay in their box and ship that.
In a strange way, the industry blowing itself up and shrinking back down might be sending us back to where I started: a few people in a room, physical or virtual, trying to build something they actually believe in, with everyone trusted to see and say what is real.
That is the loop I want to close in this next epoch. Take everything I learned in big studios and pour it back into small teams that actually want people to bring their whole lane‑breaking selves to the work.
If those teams get it right — build the real safety, trust the polymaths, let critique cross lanes — they don’t stay small forever. The studios that come out the other side of this restructuring might be built on better foundations than the ones we just lost.
In the age of AI, more people can act like polymaths: curious enough to autodidact faster, cross a few more lanes, wear more hats. I’m not speaking theoretically: I’m on a Sony PSS-AI R&D team and I collaborate with AI every day. AI is an excellent tutor: learn enough shader math to have a real conversation with rendering, enough Python to automate a boring part of your art pipeline. The old excuse of “that’s not my lane” gets thinner every year.
But polymaths and small teams only stay dangerous in the good way if they protect one more thing: how they think.
Critical thinking: question your own pillars first
A lot of people hear “critical thinking” and think “be great at poking holes in other people’s arguments.” What I mean is different: actively and honestly examining information and your own beliefs, using evidence and logic, with a real willingness to change your mind.
For me, real critical thinking starts with one-ish rule: Aim your sharpest questions at the things you believe most strongly. Don’t protect your favorite beliefs from scrutiny.
It is easy to tear into ideas you already dislike. It is much harder to stress‑test your favorites: the engine you are sure is right, the workflow you “know” is optimal, the story you tell yourself about which disciplines matter most. If you never interrogate those, you are not doing critical thinking. You are doing self‑defense. In a market this volatile, self‑defense is how you end up defending the wrong thing for too long.
On Open 3D Engine, I was certain ACES was the right call. Industry standard, the logic was sound. I couldn’t have predicted the backlash, or the arrival of AgX, which reframed the whole conversation. I also assumed engineers grasped the visual impact as well as the math. Wrong on both. Color workflows confuse everyone, including the people implementing them. It takes multiple minds across lanes to make that final output great.
- Curiosity over certainty: lead with “what am I missing?” instead of “how do I defend this?”
- Assumptions first: question your own, especially when you are the most senior person in the room.
- Evidence over ego: let data, experience, and logic weigh more than your attachment to being right.
- Humility: accept that you can be wrong even in your specialty, and that other disciplines may see things you do not.
You cannot have a culture that truly questions its own pillars if people get punished for pointing at cracks in the foundation. And you absolutely cannot have it if leaders never model “here is something I was sure about last year that I am not so sure about now.”
Great teams do not just invite critique of other people’s lanes. They create space – and obligation – to question the lane markers themselves, instead of enforcing them through lane‑policing when it feels uncomfortable.
All of that sounds abstract until the market kicks your door in. Right now, it is.
Late‑career in the age of AI (prospering from chaos)
I am writing this at a weird moment.
The games industry is going through what feels like an extinction event. The landscape evokes The Walking Dead. Whole studios are gone. Veteran teams scattered. Layoff waves rolling through year after year. AI is racing forward, grabbing attention and money, while old pipelines get squeezed and questioned.
I am late‑career in that storm. I have been doing this for decades. I have seen booms and busts, console cycles, platform shifts, acquisition cycles, and phoenix movements. This one feels different. It is not just a downturn. It is a complete restructure.
I do not want to just survive it. I want to embrace it and prosper from the chaos.
Not by pretending nothing is changing. Not by clinging to old ladders. I want to use everything I have learned (the craft, the cross-lane scars, and the leadership) and aim it at what comes next. All the things that got me in trouble in rigid systems (cross‑lane thinking, asking hard questions, refusing to play smaller than what I actually see), I wear as a badge.
AI is not going away. Layoffs will not magically reverse. The industry is going to be leaner, weirder, and more experimental for a while. That environment actually favors the kind of person I have become: cross‑disciplinary, re-skillable, able to see systems end‑to‑end.
Late‑career, for me, is not winding down. It is changing games.
That looks like:
- Treating AI as a force multiplier for small, sharp teams, not as a replacement for people who care.
- Staying brutally curious about tools, models, pipelines, and people.
- Saying yes to teams that want that kind of lane‑breaking energy, and no to ones that only want it on the slide deck.
I do not know exactly what the next epoch of games will look like. Nobody does. But I know the role I want to play: the person who walks into the chaos with eyes open, asks the hard questions, connects the disciplines, and helps build something better than what we just lost.
I’m not writing any of this as a neutral observer. I’m writing it because I want things to get better, and because I still believe they can. If any of this hits something real for you, if you want to unpack it, argue with it, pressure‑test it against your own studio, or figure out what it would look like in your context, reach out. If you want discourse, advice, or help, I will make time.
Resources and further reading
- “They work in the unmapped spaces. Between disciplines. Between what is known and what is only sensed. In the intersections that don’t officially exist because no single field has claimed them.”; If you Think in Systems and See Patterns that Others Don’t, This Is Your Moment.
- Berkeley Executive Education: Psychological Safety at Work
- PMC: Psychological Safety and Worker Wellbeing (open-access, peer-reviewed)
- Amy Edmondson: The Intelligent Failure That Led to the Discovery of Psychological Safety (Behavioral Scientist)
- Edmondson: Psychological Safety: Why It Matters (YouTube, ~10 min)
Jonathan Galloway is a polymath technical art leader who bridges art, tools, engine, and AI to help teams ship large, beautiful games with fewer headaches. A 30+ year AAA veteran with deep roots at Sony PlayStation, he specializes in technical art, real‑time rendering, Python tooling, and procedural worlds. Equal parts artist, engineer, and systems thinker, he turns messy pipelines into creative playgrounds where technology actually serves the people using it.