dasil003

If it’s truly only HR and then direct to full interview panel then I agree with you, but I’ve never worked somewhere where a technical person wasn’t involved in screening. Yes recruiters will winnow the inbound, but usually there a technical phone screen, hiring manager screen, or both.

One can find a reasonable rebuttal for anything. The point is that pattern only emerges over time—this guy hates exercise and has a knee jerk rationalization to suggest any exercise is bad.

I think most people misunderstand the relationship between business logic, architecture and headcount.

Big businesses don’t inherently require the complexity of architecture they have. There is always a path-dependent evolution and vestigial complexity proportional to how large and fast they grew.

The real purpose of large scale architecture is to scale teams much moreso than business logic. But why does headcount grow? Is it because domains require it? Sure that’s what ambitious middle managers will say, but the real reason is you have money to invest in growth (whether from revenue or from a VC). For any complex architecture there is usually a dramatically simpler one that could still move the essential bits around, it just might not support the same number of engineers delineated into different teams with narrower responsibilities.

The general headcount growth and architecture trajectory is therefore governed by business success. When we’re growing we hire and we create complex architecture to chase growth in as many directions as possible. Eventually when growth slows we have a system that is so complex it requires a lot of people just to understand and maintain—even if the headcount is longer justified those with power in the human structure will bend over backwards to justify themselves. This is where the playbook changes and a private equity (or Elon) mentality is applied to just ruthlessly cut and force the rest of the people how to keep the lights on.

I consider advances in AI and productivity orthogonal to all this. It will affect how people do their jobs, what is possible, and the economics of that activity, but the fundamental dynamics of scale and architectural complexity will remain. They’ll still hire more people to grow and look for ways to apply them.

The WYSIWYG markdown editor and wiki links are what I like about it. I’m not into the whole plugin lifestyle that seems to be a thing among the “productivity” crowd.

Having seen what happens when TPMs run things and otherwise smart engineers just do what they’re told, I can assure you this will not lead to good outcomes no matter how good AI gets.

Agreed, and ultimately it's fine because they're talking about products not technology. If these products go in a completely different direction and LLMs become obsolete the AI label will adapt just fine. Once these things hit common parlance there's no point in arguing technical specificity as 99.99% of the people using the term don't care, will never care, and language will follow their usage not the angry pedant.

Project managers shouldn't be running engineering. They are there to keep the trains running on time, not to design the track, trains and stations.

The generally accepted roles are Product decides what we need to build, Design decides how it should work from user perspective, Engineering decides how to build it at a reasonable upfront and maintenance cost. This involves a fair amount of influence, because Engineering is better equipped to describe the cost tradeoffs than any other function. Of course this comes with the responsibility of understanding the big picture and where the business wants to go. IMHO you should not be speaking to project management about code quality, you should maintain ground level quality as you go, for bigger refactoring/cleanup this needs to be presented to Product leadership (not project managers) in terms of shoring up essential product complexity so it's easier for customers to use, less support, and simpler foundation for the next wave of features. Never talk about code with non-technical stakeholders.

Funny how Steve Jobs famously derided Dropbox as "a feature not a product" and yet even after trying for decades Apple can't get that feature right.

I agree with the sentiment, but not the conclusion. Sure, numbers can be abused, just like anything else, but they provided specificity which you can then interrogate and call bullshit on. I won't necessarily fault someone for leaving off the numbers and just speaking qualitatively to the large scale system rewrite they did, but it's harder to evaluate whether such an effort was indeed warranted or was just a lateral move post-hoc rationalized by an engineer who didn't understand the original system and needed to rewrite it just to achieve that understanding. Again, if someone is satisfying the business with such efforts, more power to them. As a hiring manager, I don't want to get into a subjective evaluation of the relative engineering value of specific work at an external company that I have no first-hand context on, but I do want to know that candidates understand the highest level goals they are hired to contribute to. Metrics, however flawed, give a good entry point into such conversations.

Intellectuals and academics coming to these conclusions and talking down to the populace is a big part of what has fueled anti-intellectualism and paved the way for demagogues to take over. If your response to today's ugly political landscape is that people are stupid, then you're not helping.

Yes, and also writing for an LLM to consume is its own skill with nuances that are in flux as models and tooling improves.

I liked your point about tribalism up until you said one tribe is rational and the other not. The distribution of rational behavior does not change much tribe to tribe, it's the values that change. As soon as you say one tribe is more rational than another you're just feeding into more tribalism by insulting a whole group's intelligence.

I think the real problem is that zero friction global communication and social media has dramatically decreased the incentive to be thoughtful about anything. The winning strategy for anyone in the public eye is just to use narratives that resonate with people's existing worldview, because there is so much information out there and our civilization has become so complex that it's overwhelming to think about anything from first principles. Combine that with the dilution of local power as more and more things have gone online and global, a lot of the incentives for people to be truthful and have integrity are gone or at least dramatically diminished compared to the entirety of human history prior to the internet.

Feels a little like AWS in that regard, no?

Credo!

Your criticism of the specific details rings true, but I also liked the overall thrust of the GP which is that two common failure modes for working software engineers are either being overly academic and not efficient at practical application of that knowledge, or else being too superficial and direct about solving the immediate problem in front of them without recognizing or even being aware of the theoretical knowledge and concepts that can greatly improve their local solutions.

I think it’s fair to say those failure modes tend to disproportionately accumulate university graduates and self taught developers respectively. As long as we don’t use it as some kind of litmus test then I don’t think it hurts to call that out.

Life’s not fair in general. Personally though I have no regrets starting my career in 2001. Sure it was harder to get a job and it paid less, but it made me more willing to join a startup which gave me experience and skills that opened more leadership doors later.

By contrast I see a lot of folks with 10-15 years experience whose “normal” is an unprecedented bull market where stock is continuously up and to the right with very little correlation to their actual work. Yeah they made more money than I did, but I don’t believe their psychology is in a better place.

Either way, you can’t control for these things in life, just have to play the hand you’re dealt.

One of the beautiful things about the human experience is that there can be multiple points. The greatest works of art can be appreciated in many ways, and the viewer's perspective can add a lot of richness that goes beyond the original intent. Even in a standard white collar work environment where we want to make single points very crisply there is a real art to framing things and choosing the right words to motivate different individuals with different contexts to do the right things to row the boat in the same direction.

I think AI is useful for sifting through high volumes of data to get the gist, but I don't really count it as it's own modality. It is by definition a watered down version of the training data that produced it, it lacks the human spark that makes content worthy of attention and analysis.

I don’t know your company but this thinking doesn’t necessarily follow logically. In a large company the value of developers is not distributed evenly across people and time, and also has a strong dependency on market realities in front of them.

While it’s true that lots of companies are getting some value out of LLMs, a much larger number are using them as an excuse for layoffs they would have wanted to do anyway—LLMs are just a golden opportunity to tie in an unmitigated success narrative.

Two things are very clearly true: 1) LLMs can do a lot of things that previous computing techniques could not do and we need time to figure out how best to harness and utilize those capabilities; but also 2) there is a wide range of powerful people who have tons of incentive to ride the hype wave regardless of where things will actually land.

To the article's point—I don't think it's useful to accept the tech CEO framing and engage on their terms at all. They are mostly talking to the markets anyway. We are the ones who understand how technology works, so we're best positioned to evaluate LLMs more objectively, and we should decide our own framing.

My framing is that LLMs are just another tool in a long line of software tooling improvements. Sure, it feels sort of miraculous and perhaps threatening that LLMs can write working code so easily. But when you think of all the repetitive CRUD and business logic that has been written over the decades to address myriad permutations and subtly varying contexts of the many human organizations that are willing to pay for software to be written, it's not surprising that we could figure out how to make a giant stochastic generator that can do an adequate job generating new permutations based on the right context and prompts.

As a technologist I want to understand what LLMs can do and how they can serve my personal goals. If I don't want to use them I won't, but I also owe it to myself to understand how their capabilities evolve so I can make an informed decision. I am not going to start a crusade against them out of nostalgia or wishful thinking as I can think of nothing so futile as positioning myself in direct opposition to a massive hype tsunami.

This article glosses over what I consider to be the killer workflow of Fugitive. Specifically the ability to trace history extremely efficiently across any kind of refactoring.

The flow is: 1. :G blame to blame the current file in a locked split 2. navigate to a line, then hit o to open the commit in another split 3. search for a change of interest, then navigate to either the --- or +++ line and hit o to open the before or after version of the file in another split 4. repeat steps 1-3 as needed to drill to whatever history you need 5. ctrl-w o to kill all the splits and leave you with just the buffer of interest

whereas I'm pretty happy with git CLI for the majority of git interaction, and a graphical client like gitk for visualizing branch history, I haven't found any other history inspection technique or tool that comes close to this.

What percentage of revenue do ads and hardware contribute to the bottom line in Apple and Google? That answer will tell you more about leadership incentives than just hand-waving away the discussion based on the fact that big tech companies tend to dip their toes in a lot of pools.

It’s weird to frame Belicheck as a talent picker first. Yes, he had a lot of control but he was a coach first, not a GM. The thing that made him extraordinary was not identifying talent it was orchestrating a team system to take advantage of individual talents. Compared to other coaches that had one system and tried to fit players rigidly into it, Belicheck was master of adapting the system to the personnel. Of course he also had Brady and a lot of control on personnel, but it’s ridiculous to speak as if it was primarily his taste that made the Patriots great.

Of course there’s a survivorship bias. Everyone’s gotta eat, and the path of least resistance is to get a job and do what you’re told. Finding an alternative path is much harder of course, but in the grand scheme of things well supported by our industrial prosperity and individualist culture.

If you look around you’ll find more people doing it than you think, they just tend to be less famous than business moguls since peculiar interests are more of a niche thing but everyone is interested in material success.

What? No, caching means a specific thing: keeping a copy of data away from the source of truth, closer to where you want to read it. Caching always makes systems more complex, it never makes things simpler, and it damn sure doesn't serve as any kind of abstraction unless you're redefining what words mean to indulge your technical philosophizing.

Shades of "you need a thneed"

This article sets up a strawman by talking about someone from HR asking for an exit interview. Well yeah, you shouldn't give unfiltered feedback to HR because HR is clueless about engineering and have tenuous context at best about the majority of feedback a technical person would be inclined to give. On the other hand, there might be some legitimate feedback that A) would be relevant and B) could be helpful, but only if the organization is prepared to hear it, which you need to use your judgement on. Be warned that if you aren't regularly hearing about what the business needs, or if what you are hearing sounds like nonsense, then you are not in a position to deliver useful feedback to HR and you only have downside.

The place to really consider more direct feedback is with people that you have worked closely with. Personal relationships matter infinitely more than HR or any "official" record. If you have good relationships with your boss and/or peers, talk to them and give your thoughts, it could give you some closure and maybe even potentially improve things. Just don't let it turn into a unconstructive venting session. Ultimately working in organizations is hard and every single person can generate a laundry list of complaints, the real value is in finding a path to improve with the levers under your control. If you have credible idea about how to nudge things in the right direction, people will tend to appreciate that; if you're just looking for commiseration about how broken everything is then keep your opinion to yourself.

Uber couldn’t exist before a critical mass had a smartphone with GPS in their pocket. I don’t see what the bar is if we don’t consider it revolutionary. Anything that goes mainstream is going to eventually feel pedestrian and watered down from a tech perspective. But if you look at how most people lived and worked 30 years ago to today it’s a massive change.

Intelligence and rational thought is useful, but like any strategy it has its tradeoffs and limitations. No amount of intelligence can overcome the chaos of long time horizons, especially when we're talking about human civilization. IMHO it's reasonable to pick a long-term problem/risk and focus on solving it. But it's pure hubris to think rationality will give you anything approaching high confidence of what the biggest problems and risks actually are on a 20-50 year time horizon, let alone 200-500 years or longer.

The whole reason we even have time to think this way is because we are at the peak of an industrial civilization that has created a level of abundance that allows a lot of people a lot of time to think. But the whole situation that we live in is not stable at all, "progress" could continue, or we could hit a peak and regress. As much as we can see a lot of long-term trajectories (eg. peak oil, global warming), we really have no idea what will be the triggers and inflection points that change the social fabric in ways that are unforeseeable and quickly invalidate whatever prior assumptions all that deep thinking was resting upon. I mean 50 years ago we thought overpopulation was the biggest risk, and that thinking has completely flipped even without a major trajectory change for industrial civilization in that time.

I find it helpful to ignore the hype and the pitches, and instead focus on how AI tools are enabling better software to be made. I may personally prefer old workflows, but if they are inefficient and I cling to them, then my value as a software professional will trend to zero over time. On the other hand, the volume of software and it's importance is still continuing to increase—AI is only accelerating this trend. So understanding how software works and what AI can and can't do with it has never been of higher value.

Sure investors and CEOs want to reduce software engineering costs, but at the end of the day software is built to serve human needs, and only humans can reason and make a judgement call about whether software systems are working well or not, and because software is so precise and deterministic, there will always need to be someone who thinks like a programmer to tell the AI what to do with sufficient precision to be useful. Sure I can imagine AGI could at some point invalidate that thinking, but I believe we are very far from that point if its even possible, and even if we do reach that point we'll need massive social change or the pitchforks will be coming out from many directions.

You mean that people lie and cheat? That's always been the case. The point of honesty and character is precisely that they reflect a person's ability to value a higher good than their immediate self-interest. Let's not throw the baby out with the bathwater.

The fact that reputation has been subjected to unprecedented arms race in the face of the internet and social media doesn't fundamentally change the game, it just makes it more exhausting and overwhelming to pay attention to.