What CPython Layoffs Taught Me About the Real Value of Expertise
Posted by grandimam@reddit | Python | View on Reddit | 84 comments
The layoffs of the CPython and TypeScript compiler teams have been bothering me—not because those people weren’t brilliant, but because their roles didn’t translate into enough real-world value for the businesses that employed them.
That’s the hard truth: Even deep expertise in widely-used technologies won’t protect you if your work doesn’t drive clear, measurable business outcomes.
The tools may be critical to the ecosystem, but the companies decided that further optimizations or refinements didn’t materially affect their goals. In other words, "good enough" was good enough. This is a shift in how I think about technical depth. I used to believe that mastering internals made you indispensable. Now I see that: You’re not measured on what you understand. You’re measured on what you produce—and whether it moves the needle.
The takeaway? Build enough expertise to be productive. Go deeper only when it’s necessary for the problem at hand. Focus on outcomes over architecture, and impact over elegance. CPython is essential. But understanding CPython internals isn’t essential unless it solves a problem that matters right now.
ShepardRTC@reddit
Honestly, I lost a lot of respect for Microsoft for doing this. And doing it while the team was off to a conference. Cowardice the entire way.
qckpckt@reddit
The takeaway you should have from this is that you have absolutely no control over whether your employer will fire you or not, regardless of how valuable you are or how high quality your work is. And, that attempting to convince yourself otherwise in the manner that you seem to be is probably bad for your career and mental health.
You seem, like many people, to be clinging desperately to the belief that large organizations like Microsoft know what they’re doing and make difficult but sensible and well-informed decisions about who to lay off and when. They absolutely do not.
I have worked at or for many large multi-billion dollar organizations, and they are all equally mediocre in almost every measurable way, excluding their ability to (sometimes) make multi billions of dollars.
We live in a C- world. We do not live in a meritocracy. There are A+ people in it, but they are not in charge, because those people tend to have decency, principles and a moral code. It turns out that those are exploitable weaknesses, and as such the less scrupulous, the dull, the sycophantic and sociopathic are in charge because they simply lack the moral fiber to be disgusted at their own actions.
vulgrin@reddit
Someone wise once told me: businesses are run by C students while the work is done by the A students.
thinkscience@reddit
there is a research that shows sociopaths make good leaders !! and always remember you are a number on an excel sheet in large corps !
neddie_nardle@reddit
"We live in a C- world" is something that could be construed two ways. We live in a C minus world. True. We live in a C suite world - as in a world of managers and executives. Also true.
Amazing the number of business where those who actually contribute, those at the coalface doing the real work are utterly unimportant compared to the mediocrities ruling them. And gawd help you if you, no matter how inadvertently, somehow upset the ego of the mediocre.
russ_ferriday@reddit
Saved. Excellent summary of it all.
bliblufra@reddit
Perfectly worded. You spoke all the thoughts I had in mind
RedditSlayer2020@reddit
You just figurout capitalism. You are measured by how much revenue you generate now or in the foreseeable future. Capitalism kills true innovation and genius. Resist the ghoul class.
not_sane@reddit
What non-capitalist country do you think is worth emulating?
RedditSlayer2020@reddit
One that doesn't get terrorised by the unites states of America and their cucked puppet states
larsga@reddit
Do feel free to move to Russia. Nobody's stopping you. Enjoy!
HommeMusical@reddit
Wow, that comment would already have been a self-parody in the 1950s.
Given that Russia has now gone all in on Darwinian capitalism, it's almost incomprehensible in 2025.
larsga@reddit
This is a complete misunderstanding. There is no free capitalism in Russia. In Russia political connections determine who gets to own what business, so starting a business is highly risky, since if it is successful chances are high that it will be taken away from the founders. In fact, already over a decade ago, the most effective way to end up in prison in Russia was to start a business.
RedditSlayer2020@reddit
Russia is not a communist country.
not_sane@reddit
In principle you mean Cuba, but without sanctions? Or maybe Vietnam, but it is pretty capitalist these days.
RedditSlayer2020@reddit
Reminder: This is a python sub I just answered out of kindness.
asphias@reddit
i currently work for a government institute in a european country. the value we work for is ''value to society'', not ''profit''.
even before changing the entire system, you can go work for government or nonprofit orgs, or start one yourself.
Kindly_Climate4567@reddit
Communism even more so: everyone who drives innovation or is a genius is a potential threat to the regime.
RedditSlayer2020@reddit
Is that hearsay? I lived in a how you call it communist/socialist regime for 20 years and innovation was rewarded. Its blatant speculation on your part fueled by capitalist/imperialist propaganda.
thinkscience@reddit
they had differences, industry is going towards rust and go and investment into python felt like diminishing returns !
theacodes@reddit
I mean I'm cool with everything you said- good overall advice and such, but it is important to note that they weren't laid off because they didn't deliver clear, measurable business impact. They were laid off because of shitty leadership that focuses on short term gains that make investors happy. At a company as vast and wealthy as Microsoft, headcount is a neverending shell game that is entirely divorced from profit.
Shwayne@reddit
How can you say headcounts are divorced from profit right after saying that lay offs happened because of focus on short term gains lol
HommeMusical@reddit
The point is that these people were fired in order to juice short-term, quarterly results, but in the long term, investments in workers tools are almost always very profitable.
(I've been on the Internet for a very long time now, and yet one thing continues to be the same: comments that use the word "lol" are generally poorly thought-out.)
dparks71@reddit
I actually agree as a former F500 manager. Microsoft is so large, their services so complex and their revenue and expense streams so complicated that any headcount reduction is basically a good will gesture to the investors to say "hey, we're doing something".
People act like they do advanced analysis on reliable metrics to say "this reduction in expenses by y will affect our revenue by x resulting in z benefit for our shareholders.", but the reality is large businesses don't work like that.
They say "the shareholders are mad, what can we do to calm them down in the immediate future?" and sit in a room and kill arbitrary projects, usually by using a filter button in an excel sheet until they reach a pre-agreed upon number they feel the shareholders will be satisfied with and they won't try to force the board to do a leadership realignment.
dubious_capybara@reddit
It's not divorced from profit at all. Each manager is accountable for their team's (perceived) contribution to profitability.
BosonCollider@reddit
They claimed to pivot towards AI while firing their director of AI. I would not read too much into anything they do
dubious_capybara@reddit
Getting rid of unproductive managers doesn't contradict investing in that area.
BosonCollider@reddit
Well, I guess getting rid of senior devs mean you do lose the ability to write code that AI could not write and therefore is a kind of pivot to AI
Ashamed-Simple-8303@reddit
Perceived. Thats the keyword and it means that poltics and social skills matter way way more than technical expertise.
dubious_capybara@reddit
Right. So, not divorced from profit.
turbothy@reddit
"Perceived" doing a lot of heavy lifting here.
Funny-Recipe2953@reddit
This. Yes.
RearAdmiralP@reddit
A couple of years ago, another senior guy and I had just finished up some high profile projects, so we were free to do new things. Our boss particularly liked some of the code we had written in our respective projects, and he invited us to take that code and make it available in libraries for internal use.
We work for a hedge fund. Everything is about profit and loss (PnL). While we were working on the library project, my colleague made an observation: "We're getting a little bit far from the PnL." I got the implication. If the guys doing the trading are making the PnL, and the teams providing the tools they use are one step away from PnL, and the teams providing support for them are two steps away, etc., we, with our little library project, were not even in the same room as the PnL. It's not the best position to be in for career purposes.
These days, I work on a trading critical system, and he's working on a project with a major institutional push behind it, and we both feel secure in our positions.
Anyway, those Microsoft guys fell into the trap that we avoided. They got too far from the PnL, so they were disposable when it came time to make cuts.
HommeMusical@reddit
Quarterly P&L. That's why these short-sighted, stupid decision continue. Having better tools almost always pays off even in the medium-term.
TraceyRobn@reddit
In the old days, big companies had RnD departments. They researched new technologies, things that wouldn't make money for a few years. This sort of research gave us information theory, the transistor, C, and Unix.
These days the time horizon is the next quarter's results and the latest buzzword (AI at the moment).
We've eaten our seed corn.
vexatious-big@reddit
Well said. There's a good article on exactly this here:
https://1517.substack.com/p/why-bell-labs-worked
Diligent_Day8158@reddit
But what’s the point if all we do is sustain what works? That’s not how real growth happens right? American economy has been getting ruthlessly slashing R&D for a long time now and tech is catching up to that theme.
pythosynthesis@reddit
I was laid off in the past for doing exactly what my boss asked me to do, which was to implement a new model for the desk that would have been used down the line to support the business initiatives, but the cuts came "now", not "down the line". Guess how much PnL I had to show for my work. My boss even apologized a few years later.
papawish@reddit
Last year my team has seen layoffs. Almost all teams affected.
My team has had 2 net hires lol. We are the golden goose.
Follow the money and stay tied to it
yangzhou1993@reddit
Job security is dead in the age of AI.
ufos1111@reddit
Even microsoft thinks javascript is better than typescript these days lol
grandimam@reddit (OP)
Just curious, how did you arrive at that conclusion?
Jubijub@reddit
In one sentence : always understand the link between your work and where you paycheck comes from (ie where the company earns the money that pays you). The more convoluted / unclear that link is, the more you are at risk if a down cycle or vicious cost cutting exercise occurs.
No_Challenge_9867@reddit
Just chill. They must have by now have got absorbed in different roles or will get soon in Microsoft itself.
kylotan@reddit
The problem with that is, while it's very good career advice, it's not ideal for the development of technology that requires deep understanding and solid foundations.
A lot of the problem we have in computing, from critical ones like security, to less critical but widespread ones like poor performance, come from people focusing on shipping first and correctness later.
There needs to be a better way.
Wh00ster@reddit
This is great, but also like basic knowledge 101.
It’s why research does not pay well unless it’s been clearly monetized.
DadAndDominant@reddit
That has always been the truth. Meritocracy is a lie, because there is no merit; your value is only measured as the difference betweeen the costs to hire you (input) and costs of what you produce (output).
Real late stage capitalism just isn't meant for humans to participate in, just a selected few to reap all the benefits. It is commimg faster and faster.
floriv1999@reddit
While what you said is generally true, there are a few other things to consider. Mainly short vs. long term profits and the size of the organization. For optimization work larger orgs benefit a lot more. If e.g. somebody squeezes out half a percent in performance this means millions of dollars saved for an org like Microsoft. You can easily calculate it if you have their internal data on it, so it is a measurable profit. Therefore it makes a direct impact. But it might take some time to add up, making it more of a medium to long term investment. In a startup having somebody on the CPython or TS team would not make so much sense, because they have less cost associated to running these languages in production, so the upfront cost of paying them their salary is not worth it. But somebody else can always optimize it right? As the salary is pretty negligible compared to the savings you are better of being in control of what is implemented when, giving you the most gains and also the highest (soft) power in the project, which can be useful for many things. Then there is PR. Companies like MS have giant PR teams for more or less appealing PR campaigns. I think doing stuff like developing TS benefits the perception of MS in the technical community. It is relatively cheap PR for them. Remember Advertising etc. are really expensive.
But why do they still lay them off? Because managers are also humans and doing layoffs is tending and cool now. We are way less rational then we always think we are. Also internal politics in large organizations are a whole different beast. It could be the case that the teams are paid for by a different part of the organization then then one who benefits in the end. Managers mostly decide what's best for them and their part of the organization, compared to the organization as a whole long term. So paying for somebody who benefits another part of the org makes no sense for how their side of the business looks on paper. And maybe in the end some decisions are just stupid and made by people who rose in the hierarchy without being very competent. That also happens a lot on large organizations.
mm_reads@reddit
Or just stop providing technology for profit-only capitalism.
Better yet, start gunking up AI.
AI is just a tool. Until it is a tool FOR people, and not a dehumanizing & destructive environmental force, it's just a weapon to expedite human misery.
scilente@reddit
This is AI generated, yeah?
florinandrei@reddit
TLDR: Not making the rich people even more rich? Chop-chop your head.
fgiohariohgorg@reddit
That's Capitalism: you get what you deserve, as long as you're convenient for me. Sad that people are dumb enough to just care for $
SEC_INTERN@reddit
Obviously ChatGPT generated post.
grandimam@reddit (OP)
Not entirely. But I did use Grammar correction tools for help.
btoor11@reddit
My man, you’re either a karma farming bot or got your whole spiel written by a chatbot.
Your opinions do not matter. Because I’d be damned if I take career advice from someone who doesn’t even have the skills to write a coherent opinion without a handhold from Ai.
grandimam@reddit (OP)
English is not my primary language. So, I did use Grammarly freemium for correcting the grammar mistakes. Will do it better next time. Thanks
fredandlunchbox@reddit
This isn’t always true.
In times of abundance focused on growth, research, optimization, and structure gain importance.
nukem996@reddit
I know some very good engineers that went product because it's better long term to follow the money. Once you know maximizing profit is the only goal it can make decisions much easier.
grandimam@reddit (OP)
When you said product? You mean Product Management or Product Engineer?
Kindly_Climate4567@reddit
Working directly on a money making product, not in a support role that is a cost centre.
Sherpaman78@reddit
You are right, and everyone is right, but it is not a natural law, it is a law of Capitalism.
Your value, in Capitalism, is not measured by your expertise or knowledge, but by the value a company can extract from your labor.
Remove the "Capitalism" from the picture and it is not necessarily true anymore.
Cytokine_storm@reddit
Nah, this is true for any society with resource scarcity.
BogdanPradatu@reddit
Just that Microsoft doesn't have a resource scarcity problem.
this_is_a_long_nickn@reddit
A natural one? Not at all.
An artificial “shareholder value at any cost” one? Absolutely yes.
Shensy-@reddit
Which is caused not by any real scarcity, but an artificial construct of...capitalism. Look at the pretty circle.
UltraPoci@reddit
The takeaway is: fuck Capitalism.
numerical_panda@reddit
Buying impact using technical debt.
charmquark8@reddit
The takeaway: unfettered capitalism is shit.
jake_morrison@reddit
It’s not about business value. Managers would have more, relatively junior/average developers than fewer, more highly skilled developers. This lets them build a fiefdom with more people/budget. And they want cheaper, more disposable developers with less leverage.
float34@reddit
Yeah right, then these fuckers ask you to go "extra mile" for nothing, to prove your dedication.
tthrivi@reddit
This is where innovation dies. Fed is pulling away. Companies don’t want to fund long term projects. Decline of the US as a leader of innovation and technologies.
Fabulous-Possible758@reddit
Yep, people are just resources to be allocated by a company and most of the time the people making the decisions will use apparent revenue as a qualifying factor when cutting costs. That’s… every job in corporate America.
dada-engineer@reddit
On the other hand these people have such deep expertise that they are likely to easily find a job. Not sure if this is true when you are scratching the surface of all the things
edeltoaster@reddit
A hard but reasonable truth.
BogdanPradatu@reddit
The hard, but reasonable truth is that no matter how good you are, how hard you work, how passionate you are, all it takes is a few percent drop in stock price and you're gone.
I bet there are less capable employees in Microsoft than these guys, that haven't been layed off. They could have fired some other guys and replace them with these experts, but they didn't.
The python guys were hired to work on CPython, that was their job. If Microsoft felt like it doesn't bring value to them anymore, they could have repurpuse them, but they didn't.
Most people don't get to choose on what project they work on. It's just chance.
grandimam@reddit (OP)
Yes.
Empanatacion@reddit
Apples and oranges. Microsoft was paying them to help everybody, not just Microsoft, so they didn't feel like they were spending on something giving THEM value.
But the bigger reason is that they don't give up anything concrete by firing them. CPython will get picked back up by someone, Microsoft will benefit from it, and they won't be paying.
BirdTurglere@reddit
Those people can probably walk into another job though. And with inflation being the way it has been they’ll probably get a pay boost doing so.
Not trying to down play layoffs. Some of those people might have wanted to settle down etc and that’s definitely unfortunate. Microsoft’s loss for always chasing hype.
But if there were only 100 python jobs left on the planet they’d be in a good spot for them.
Just being a grunt doesn’t give you that security.
bestjaegerpilot@reddit
i like working on tooling but those are the first to go. The issue is companies focus on the short term. In this particular example, they're not measuring how TS has been making their org more productive.
If you squint, this also marks the end of TS. It's gotten so complex only a well-funded company will maintain it but that well-funded company just said the hell with it.
It's gonna be a wild west w/ peeps trying to make Rust/Go/C a thing via web assembly.
SailingToOrbis@reddit
Yeah the business executives never give a shit about how good the technology is, but all they care is how much money that you can make out of it.
ElderberryPrevious45@reddit
In other words : In life it matters more what you do than what you dream of. Also, some dreams are wiser to leave that way, just dreams, because you never know all.
t1x07@reddit
Well said, I think it's one of the hardest things for programmers to actually focus on the value proposition of their work because it's so easy to get lost in optimising and tweaking things. There's a reason why there's so many memes about programmers spending hours to automated tasks that take minutes
Ok-Willow-2810@reddit
I don’t disagree with what you said! Just being very, very knowledgeable in a highly specific domain doesn’t guarantee that all work in that area is worthwhile. Work needs to bring value to a company. However, I’m also a little unsure of what the longer term impact of this sort of thing is. I for one really appreciate open source and community works and I think it makes the ecosystem better. Will this noticeably shrink the ecosystem over the next 5 years? If it does, would that have a negative impact on future market share? I don’t know how to answer that question and there’s many other factors and it’s really tough to measure. However, it also makes me wonder if things that are easier to measure and point to clear value are always the best strategy? It could be that some more complicated things to explain or see the impact of have a lot of value, but it’s simply tough to explain. I’d caution against assuming something is tough to explain or too specific doesn’t bring value. That being said, I have no way easy way to know or measure whether these layoffs fall into the category I explained and how it truly affects profitability for Microsoft.
I see it as Microsoft maybe placing less value on helping the open source community, which I personally don’t love. Maybe it won’t necessarily be bad for the open source community though in the long run? Maybe it will. I don’t know how to know that, but I like open source and I don’t like seeing layoffs. Wish the economy was simply undeniably booming!
pbecotte@reddit
Can't help but wonder if open source is on the way out. Tragedy of the commons and all that- the fundamental technologies it's more cost effective to hope someone else builds it for you.