Large tech companies don't need heroes

Large tech companies operate through complex systems of processes and incentives that determine outcomes, not individual heroics. These systems evolved over time rather than being designed, creating inefficiencies that trigger an uncomfortable reaction in engineers who are internally compelled to fix broken systems. Past a certain point, fixing these inefficiencies instead of working on your actual projects will get you punished, not rewarded. Doing it anyway requires sacrificing your career progression - becoming a hero.

The problem is that heroism doesn't actually benefit the company long-term. Your efforts in one corner of a doomed product like Google Wave won't save it. Even poorly-run teams can succeed if they're targeting a niche the company supports. Worse, heroism actively prevents real change by insulating the company from the consequences of its broken systems. Far better to let the company be punished for its failings so it can slowly adjust or be replaced.

What heroism does enable is exploitation. Product managers keep mental lists of engineers they can convince to do extra work that benefits the PM but not the engineer. During launches, organizations wage cold wars trying to extract help from each other's engineers while guarding their own resources. Managers let engineers spend all their time on glue work because it makes the manager's job easier, then punish that engineer at promotion time for not doing their real work.

The key is paying attention to actual rewards - promotions, bonuses, raises. These show what the company really values. Predators don't control those levers, so they appeal to your internal compulsion to be useful instead. A background level of inefficiency is just the price large companies pay for scale and legibility. The more you can live with it, the more energy you have to use tactically for your own benefit. Unless the work is explicitly rewarded, your heroic efforts only delay the point where the company has to fix its processes.