In Part 2, we looked at the structural pattern: how specialist systems create bias, and how it doesn’t stay inside any one discipline. Any silo can start believing its lens is the only real one. The problem is the architecture, not the people. This part is a little disjointed, but it is about the environment those people work in: what fake safety actually looks like, how stewardship and invention require different kinds of leadership, and what building a real culture of psychological safety actually takes… not just putting it on a slide.
Fear with better marketing is still fear. That’s the thread through all of it.
When “safety” turns into theater
There is another flavor of this that shows up in big‑tech and big‑brand environments.
On paper, the story is all about being a “world‑class workplace,” hitting diversity goals, and building an inclusive culture. The slide decks look great. There are hiring targets, internal campaigns, lots of messaging. And to be clear: fixing representation matters, and diversity and inclusion are a feature, not a bug. Some of the strongest people I’ve worked with only got in because someone finally took goals like these seriously. That part is real. The problem isn’t diversity or representation; it’s when we stop insisting on merit and standards as well, quietly waive the bar for optics, and then let “diversity” take the blame when those shortcuts blow up.
I’ve seen it happen. A highly specialized role with a well-understood bar (portfolio review, design calibration, technical calibration, peer assessment) suddenly gets an exception because a candidate ticks an important box for a leader or hiring manager. A portfolio doesn’t matter this time. The design review will slow down hiring. We can skip the technical screening. The justification is framed as progress: “we need more people like this.” When it doesn’t work out, the post-mortem isn’t “we broke our own process.” It’s a quiet eye-roll about “hiring for diversity” from that same leader. No accountability for the decision; everyone else loses in multiple ways.
For underrepresented folks who did clear the bar, this is poison: it makes them wonder whether every side-eye is about their work or about someone else’s shortcut. The individual is set up to fail, the team’s trust in the standards erodes, and people who genuinely did clear it start wondering who’s silently questioning their legitimacy.
That’s the failure mode I’m talking about. Not “merit versus diversity” as a fake binary, but leadership using identity to justify inconsistent standards, then letting existing personnel take on the tarnish and marginalized groups carry the reputational damage instead of owning the call.
The problem is the shellac.
As a professional your artifacts and assessments are required, and you expect to take hits and roll with punches along the way. In big tech sometimes it feels like you’re in the octagon. Your own docs are expected to be brutally combed over, commented on, and objectively reviewed for accuracy and value. But with certain peers, you’re expected to walk on eggshells, deliver meaningful feedback indirectly, behind closed doors and through middle‑people. That is not safety. That’s coddling dressed up as care.
Teams or individuals can ship obviously weak work and, in a more public forum, it gets coated in a glossy layer of praise: “So proud of this team,” “Incredible effort,” “Amazing outcome,” “Great work.” The real conversation happens in private: here is what broke, here is who isn’t delivering, here is where the bar actually is. This isn’t about any one person’s merit; it’s about how organizational incentives decide which truths get spoken out loud and which get buried.
I grew up and cut my teeth in cultures where excellence was the North Star. The lane was “best game we can make,” full stop. We could say hard things out loud. People bristled, arguments happened, but the assumption was that we were adults and the work mattered. Thick skin was part of the job, not because cruelty was the goal, but because honesty was.
It is a bit cringe by today’s standards, but a leader who has also made some of the most highly rated games in the universe once pseudo-apologised to me “Jonny, if we are not at each other’s throats once in a while, one of us isn’t pushing hard enough.” That stuck, it was meaningful and actionable feedback. It is not tone deaf at all, quite the opposite.
In the softer, brand‑driven version, a different pattern appears:
- Public story: everything is great, everyone is killing it, the employer brand stays pristine.
- Private story: a small circle tells the truth quietly, where it feels “safe” to admit something sucks.
That is not psychological safety. That is fear with better marketing.
The goal is to build a studio culture and player community that lasts forever. Although we can’t carry on with toxic grind and crunch culture, and we have to look that in the face… “Optimize for the player experience, and the hits will follow.” is a better driving force but this itself is a relentless search and requires truth seeking.
Real psychological safety is permission for candor. It is the ability to say the uncomfortable thing in the room where it matters, push for the highest standards, and not be punished for it. What you see instead is a kind of organizational “ruinous empathy“: leaders are so afraid of bruising anyone or scratching the brand that they stop giving clear, direct feedback.
You cannot optimize for a flawless public culture story and quietly de‑optimize for truth, merit, and excellence without paying for it. The price is trust.
People are not stupid. They know when output is weak. When messaging is flimsy. They know when someone is struggling. When the official story is, “everything is amazing,” and the unofficial story is, “we’re in trouble, but don’t say it out loud,” you train everyone to play along.
You create a culture where:
- Saying the honest but hard thing in public is risky.
- Saying nothing and nodding along is safe.
Call it what you want, but it is not inclusive and it is not safe. It is just another form of lane‑policing: “Your job is to support the narrative, not to say what you actually see.”
The places that sometimes feel harsh on the surface – where someone will look you in the eye and say “this isn’t good enough yet” – are often the ones that are truly safe. The kindness is not in the sugarcoating. It is in being willing to tell the truth while still having each other’s backs. Delivering objective feedback and criticism is a skill; how it’s delivered matters. But perfect is the enemy of good: people need psychological safety to use their voice and offer real feedback before you can help them hone it. That, too, is a form of stewardship, here is another …
Stewardship vs invention
There’s a big difference between stewarding something beloved and inventing something new.
Some of the best work I have seen in this industry has been faithful remakes and remasters of classic games. The praise is almost always about stewardship: preserve what made the original special, modernize visuals, controls, and feel.
That is a specific kind of problem:
- The core creative identity already exists: tone, mechanics, pacing, story, fantasy.
- The audience contract is known: fans will tell you what is sacred and what is negotiable.
- The question is not “what is this game?” but “how do we honor this game now?”
It is still hard, but the ambiguity is narrower. If you corral ego and align on what is sacred versus what is fair game, the target is clear. You are aiming for “this feels like I remember it, but better,” not “this is some brand-new thing nobody has a mental model for yet.”
New IP is a different universe.
There is no proven fantasy. No fan contract. No existing shape to protect. Every pillar… mechanics, world, story, camera, systems… is subject to change, and most of your strongest early beliefs will be wrong in at least one important way. The risk profile is not “don’t screw up this beloved thing.” It is “can we even find something that deserves to be beloved?”
In a remake, a top‑down “we already know what this is, just execute” culture can survive, sometimes even thrive, because the original game quietly does a lot of the heavy lifting. In a new IP, that same culture is a trap:
- If dissent is unwelcome, you cling to bad assumptions too long.
- If lanes are rigid, nobody challenges the core fantasy or scope until it is baked into the schedule.
- If leadership treats the game as a fixed vision instead of a hypothesis, you only discover you were wrong when it is very expensive.
This is where psychological safety and critical thinking go from “nice ideas” to survival traits. In a risky new IP, you need people from every discipline asking, “What if this pillar is wrong?”, “What if this workflow will not scale?”, “What if players do not want this?” And you need leadership that can hear that without treating it as a personal attack.
Remakes reward ego control and craftsmanship. New IP punishes ego and rewards environments where it is safe to question even the things the director believes most strongly.
Stewards of culture and safety
If you’re a lead, director, or manager, you’re not just shipping features. You’re the steward of the culture around them. You decide how much voice people really have, whose discomfort counts as data, and whether psychological safety is real or just a phrase in a deck.
Most people don’t need to win every argument. They need to feel heard, respected, and taken into account. Good leaders listen and ask questions. When people feel like their perspective was genuinely considered, they can disagree and still commit. When they don’t… when their input disappears into a void, or gets swatted away with lane-policing… they get brittle, skeptical, and eventually disengaged.
Good leadership doesn’t mean crowdsourcing every decision. Leaders still make the call. The difference is in how they get there: do they actively pull in voices from across art, design, engineering, QA, production, community? Do they ask real questions before they decide? Do they make the reasoning visible so people understand the tradeoffs, even when they don’t love the outcome? Do they listen to the signals, especially when those signals show up as discomfort?
In my experience, if you do that, most people will absolutely toe the line. They’ll help you make it happen. The decision is made, the ambiguity is gone, the direction is clear… and because they had a voice on the way there, they’ll put their shoulder into the result instead of quietly sandbagging it.
That’s stewardship at the level that actually matters. Not about being perfect, or never hurting anyone’s feelings. About building a team where telling the truth is safer than staying quiet, where merit and standards are applied consistently, and where people can say “I disagree, but I’m in” without feeling like they just volunteered for exile.
The same restructuring breaking the industry is cracking open the old molds. Which sounds more hopeful than I probably mean it. Part 4 is about what comes next: polymaths, small sharp teams, and what late-career actually looks like when you decide to lean into the chaos instead of waiting it out.