I have been fired multiple times in my life. Each time it has been because I violated a fundamental rule of power. Where I had stayed employed when others were cut it was also due to "observation of the laws" of power. This is not to say I understood this at the time, but to observe that "this time is not different".

The first "real job" I got out of college was testing calculators for Texas Instruments. I subcontracted there for about 4 years, and was one of the few who survived a ruthless layoff associated with the 07/08 panic. This was a very close run thing. There was one day in which I was fired and re-hired in the same day.

It is clear in retrospect that the reason I stuck around was due to being better at finding issues than all my peers. I had by that time found a number of critical issues with the multi-line scientifics by mapping out the memory pages and watching for stomped flags. Nobody else testing the products at the time came close to understanding the hardware at this level, making me indispensable.

Which is to say I focused like a good protestant work ethic boy on laws #9 and #11. Demonstrate, don't explicate. Keep others dependent on you to achieve freedom. I keep going back to this over my career, as it worked.

I also learned law #13 "Only appeal to self-interest" when it came to seeking promotion and favor from management.  I found quickly that "job descriptions" were universally meaningless and the only important thing was delivering on stuff your manager was emotionally invested in.

This was about two years before I got fired. I went on to do more things for the firm which nobody else understood, such as solving a data encoding issue with archival documents and porting the TI-8X emulator to linux. I had made a good number of friends and was well liked at the firm.

Nevertheless, this made me a bit too comfortable. I was also still a pretty naive young man at the time, and actually believed upper management would appreciate serious criticism. This is of course not the case, and they see it as an affront and out of place. To do this is to violate rules of power #1 and #19, "Don't outshine the master" and "Don't offend the wrong people". Like my victories this has also bitten me more than once.

Interestingly enough a couple of months after my ouster, I got an offer to work on the programming of the color TI-84 from one of the programmers there I had a good relationship with. Apparently the criticisms which I had of management were quite timely and the issues I had brought up promptly blown up in their face like backdraft. As such, there was no resistance to my return as all oxen gored were now out of the picture.

I had taken a job with cPanel by then though, a firm which I would spend 8 years at. I also rapidly rose to a position of indispensability in the QA organization there, but took a brief hiatus to work with my cousin at his startup HailStrike. In retrospect this should have been an obvious violation of Power law #10 "avoid the unhappy and unlucky". The company was a reject bin in many ways.

Nevertheless due to my upbringing which had turned me into the stereotypical "nice guy" who immolates himself to keep others warm, I did a lot of good work there. I built a new product from the ground up and re-wrote the existing one to not have horrible projection bugs and awful performance. That said, nothing could save that firm, as my cousin and his partner hired a con-man to run the firm thanks to their lack of self-confidence. After about a year and a successful funding round, the co-founders went on a month long vacation and returned to find the place looted.

In that time and in the aftermath I basically kept tech end of the shop going single-handedly for minimum wage. After about 6 months of this I cut bait and returned to cPanel, being close to "zeroed out" financially. All I got for the trouble was some worthless stock in a firm which languishes to this day.

Meanwhile cPanel's QA department hadn't changed much from where I had left it. They were eager to make some forward progress and remembered my impact. So my departure at least had the positive effect of resulting in a big raise. Law #16 "Use absence to increase respect and honor" in action.

For the next 5 or so years I became the most senior man in the department. I made a number of tools without which the department couldn't do their jobs. I also was #1 across the board in test execution and bug filing metrics.

So far, so good. I also cultivated better options to effect a promotion and significant raise in my last two years, but this would prove my undoing. Much ink has been spilled about how it's always better to take the other offer (much of which I had read!) but I was emotionally invested in the firm after 6 years and accepted the counteroffer.

I was now writing product code rather than automation for our QA workflow. Similar to in my prior role, I quickly rose to the top 5 bug fixers and committers at the firm. This, however is not the same as indispensability.

It turns out that you need to work on products that matter if you want to stick around. At TI, I worked on cash cows, and these things will forever need their indispensable people. However at cPanel they had a monoproduct which was itself an aglommeration of various sub-products some of which were important and not. The teams I was put on were rarely working on anything the customer was particularly interested in.

To be fair, this is the case across much of the organization. For years only the CEO's team was the one working on anything relevant to customers. The rest of the firm was run autonomously by the middle management and fell victim to the principal-agent problems that entails.

As a middle manager, to aggrandize yourself you generally want to weed out the indispensable and maximize your headcounts. This is generally accomplished by two means:

1. allowing the indispensable to silo and thus violate rule #18 "do not isolate yourself".
2. making sure you don't take any risks you can be blamed for, and blaming failures on lack of manpower

The consequence of 2) is that nothing of consequence is worked on, meaning that no new person will ever achieve true indispensability. We had a core cabal of old developers most of whom were siloed (and gradually being pushed away), or had mentally checked out themselves due to working on unimportant tasks. I was obviously not among either as I "still cared" rather than being checked out (and thus not perceived as threatening).

It also didn't hurt that up to that point I used programming as a superpower in a field that traditionally has little power in development organizations (QA), or worked on cash cows (which are rarely influential).  The powerful never feel necessitous, and when push comes to shove they'll throw their best overboard.  I had unwittingly aligned with weak factions until this point, which is a necessary (but not sufficient) condition for becoming indispensable.  Now that I was a part of the engineering department which dominated the company, there could be no allies to protect my independence.

I was eventually assigned a new team and manager which I should in retrospect have realized was a trap. I had built quite the reputation for independence while I was there (which is normal with the indispensable) and clashed multiple times with this manager. Enough things piled up over time which were not explicitly breaking the rules but did not signal submission that he formed a negative opinion of me. I'd been making a number of other changes in my life at the time which were bringing refreshing youthful joy, so I suppose it is not surprising I returned to the indiscretions from my youth which tripped me up at TI.

All it took from there was a minor dispute which could easily have been resolved peacefully being escalated in bad faith.  Some of this was simply because I fought the situation at all.  Bosses like to feel like the "cop" in the relationship in these situations, and we all know how cops feel about anyone who doesn't instantly surrender, grovel and degrade themselves for daring to attract their ire. This is why rule of power #22 is a thing.  Surrender is the best option in such situations where you are already "caught up", as bosses think any benevolence they show from that point is a thumb-screw they can use on demand. Obviously you would prefer these thumbscrews not be used, so the tactic is to buy time and enough freedom of action to get out of there.

Rule #42 also comes into play, "Strike the Shepard and the sheep scatter".  Even if you are in the right, management cannot tolerate defiance spreading.  It's simply inviting further attack.  While this is effective at keeping management powerful, it also has the effect of entrenching whatever errors they are engaged in.

At the end of the day the question that must be asked is "would you rather be happy, or right?"  Being emotionally invested in the firm you work for and your role in it means it must "do right" in order for you to be happy.  This is a recipe for disaster, as everyone's emotional needs from the firm differ and become guaranteed to clash past Dunbar's number.  This is why Power law #20 is a thing: "commit to no one".

This desire to have a useful culture at a company and a good "mission" is power law #27, "Use people's need to believe to create a cult-like following".  While you can't hate the player for "playing the game", it is straightforward to realize that there are a great deal better things out there to direct your belief and worship towards than a corporation.  All the senior developers I've known who were checked out totally about the firm had the right idea all along.  The company can want a certain culture all it wants and even go to great lengths to inculcate it, but it simply can't work past a certain scale.  You have to insulate yourself from this and resist getting sheep-dipped into their hyperreality if you want to remain happy.  Focus instead on doing the things that give you power over your situation, which is real freedom.

The shock of being removed from a place I'd been 8 years with a number of good friends took a while to absorb, but it's pretty clear where I steered wrongly. I should have learned the lesson of the Count of Carmagnola. You can't be a star when what they need is a cog.