I think root problem here is trying to estimate the net present value of work that may take several quarters to play out. This forces everyone to be fairly conservative in their estimates and generally results in institutional high time prefrence.
Idealy, comp would recognize and compensate the concrete/realized impact rather than expetcted impact but doing so would require that retroactive compensation be dependable implying either payout independent of present employmet status or long average tenures in lieu.
In this enviroment, having people reguarly & annonymously assess/attribute the extent to which their work was enabled/assisted by their peers would be more straight forward.
Transmeta lived on in Nvidia's Project Denver but Denver was optimized for x86 and the Intel settlement precluded that. It ended up being too buggy/inefficient to compete in the market and effectively abandoned after the second generation.
Starlink was originally a google idea, originally a sister project to Loon, the founder quit and brought his team to SpaceX and then got fired by Elon. Everyone who repeats this story to me says it was because "Elon doesn't suffer fools" but YMMV.
My experience has been the opposite - older googlers want to work from Tahoe/Vagas. Younger googlers want to meet other googlerrs or just travel and work a few months s year.
I agree tbh but it feels like the current environment doesn't reward risk taking and heads down work, especially with diffuse or longer term payoffs, and our AI efforts are still being retardedby safetiests/activists.
Rest & vest types are still doing fine, it's the passionate/care-a-lots that are increasingly burning out and disengaging or leaving.
Yeah I get the impression that people don't take as much pride in their work for its own sake anymore - more assembly line worker than master craftsman.