- More power to you - I easily spent weeks at a times sifting through resumes and interviewing, and even once someone was hired making sure they were onboarding well and feeling well-supported and etc.
By the time everything was humming along we were either raising a new round (and thus hiring more), or someone was leaving and we had to refill that role.
- > This is not really a good analogy. Tech is different.
Why? Seriously: Give me a convincing reason why tech is different from every other field, where this happens regularly.
> I am not saying you should sit down and write code.
But that's the whole premise of this conversation. It's entirely possible to understand something deeply without doing the thing yourself.
It's entirely possible for a CTO to deeply understand technology without writing any code themselves, opening up a terminal, tinkering with anything, or even what individual contributors are doing day-to-day. I would actually say that's the hallmark of a good CTO.
- Oh, I see now.
- I don't think this is a really convincing argument: There are plenty of leaders who haven't "done the job" in decades, and we don't question that. It's incredibly common in professional sports, for example.
Mike McCarthy hasn't played a down of American football in 40 years, and never played at a very high level. But we don't question his ability to get others to perform complex motions.
- Most of those engineers, outside of ones who have extremely specialized knowledge or skills, are essentially managing others still just without a direct reporting chain.
The commit log for most of these high-level engineers is extremely sparse. They're spending most of their time writing documents or influencing orgs, not writing code.
- > it’s not what you do day to day.
I don't know about that. I was a "CTO" for a small (10-person) and a slightly larger (around 100-person) VC-backed startup. Hiring was always top of mind at both places. Not even "people management", just hiring alone. I'm not saying this universal, but when a company is expected to scale rapidly (as is often the case with venture-backed firms), managing people can easily consume your entire workday even at a relatively small size company.
Of course, I'm not saying that's everyone's experience. There are obviously lots of reasons that dynamic might be different: For example if you're not a VC-backed company, or a CTO who's a world-renowned technical expert in a particular field (nobody's bringing Ilya on board for his ability to hire).
But it's very, very easy for people management to be a day-to-day thing and I don't think it's a waste of time compared to direct contributions.
- I don’t think we do agree, but fair enough. I’ll “disagree and commit”.
- Sure, but these are all true of middle-class employees as well:
1) Many middle-class families rent and their landlords aren’t necessarily any more understanding.
2) Not to be too political, but many middle-class employees don’t enjoy a friendly relationship with police either and similarly can easily “fail”.
If your argument is that being wealthy affords you a lot of leeway to fail in life, I mostly agree (though again, there are plenty of minority groups who would disagree that wealth always affords that privilege), but “middle class” encompasses a very wide swath of people which this doesn’t apply to. Many middle-class employees in the US are a paycheck or two away from being pretty destitute.
Maybe you meant “professional” or “upper class” instead?
- Most of the examples you gave seem focused on life outside of work aside from the last one, so I’m curious which of them you’d say don’t also apply to lower-income jobs. There are lots of ways for middle-class people to “fail” too outside of work.
Personally, I worked in food service for a decade (mostly as a line cook of some sort) and most of these rules still applied, maybe to a slightly lesser degree.
Even with dishwashing, if you have some way of dishwashing that halves utility costs, someone would listen to you.
The answer might also be “who cares, get back to work” but that’s also true of a lot of middle-class employment. Your manager won’t give a shit if you think the expense reporting system sucks. Amazon’s famous for “disagree and commit” which is just a corporate way of saying the same thing.
- I'm not of the "LLMs will replace all software developers within a year" mindset, but this critique feels a bit overstated.
The challenge of navigating rapidly changing or poorly documented code isn’t new: It’s been a constant at every company I’ve worked with. At larger organizations the sheer volume of code, often written by adjacent teams, will outpace your ability to fully understand it. Smaller companies tend to iterate so quickly (and experience so much turnover) that code written two weeks ago might already be unrecognizable, if the original author is even still around after those two weeks!
The old adage still applies: the ability to read code is more crucial than the ability to write it. LLMs just amplify that dynamic. The only real difference is that you should assume the author is gone the moment the code lands. The author is ephemeral, or they went on PTO/quit immediately afterward: Whatever makes you more comfortable.
- Which of the seven OP referred to are OpenAI or Anthropic?
- People are overlooking the fact that Twitter’s situation wasn’t exactly great even before it was taken private.
If it's in a stronger financial position today, it's almost entirely because of the merger with xAI.
- > Only seeing growth when Musk does something stupid, and most of the new users not sticking around are strong signals it doesn’t have long-term value. Bluesky is the rebound social network.
One thing that often gets overlooked is that Twitter itself was on pretty shaky financial ground (and likely still is, though being private now makes that harder to know). Even if Bluesky managed to absorb the entirety of Twitter’s user base, it’s still unclear whether that translates into a strong business model.
Yes: Plenty of criticism has been aimed at how Twitter was run, and maybe Bluesky is managed more effectively right now, but there's no evidence to suggest Bluesky would be run significantly better at that scale.
- > pragma pointer_safety strong" which would force the compiler to only accept the use of smart pointers
You’d possibly just be trading one problem for another though - ask anyone who’s had to debug a shared ownership issue.
- > I don't think it's too late for someone to fork both C++ and Clang and make something that's actually a good synthesis of the old and the new.
People have tried variants of this already: Carbon, for example. I don’t think anyone outside of Google uses it, though, and even within Google I suspect it’s dwarfed by regular C++.
I don’t think C++ will become irrelevant for a long time. Recent standards have added some cool new features (like std::expected), and personally I feel like the language is better than ever (a biased opinion obviously).
Memory management is still a huge elephant in the room, but I don’t think it’s becoming irrelevant.
- This is an odd thing to say given the events of the past few days and even last couple years:
“Nor 1960's America, where Mr. Oswald could buy a rifle and 4X scope by mail order, then get a job in a convenient book depository”
Your broader argument still holds that those in power often don’t tend to view isolated threats from the public as truly existential threats (until they do, like we just saw in Nepal).
But it’s a bit hard to agree that even in America things are truly that much different.
- Social media has proven to be quite an effective tool for mobilizing protests and beyond. I get how the short-sighted might see it as a tactical move to "cripple logistics" by banning social media.
But, the reason I call it short-sighted is exactly what you said: Removing those earlier pressure-release valves doesn’t solve the underlying issue at all and just increases the risk of a more volatile outcome.
- Equities markets are largely driven by institutional investors, save for some notable exceptions ("meme stocks").
Unless the theory is that institutional investors are doing the same, it's not that surprising.
- > I still think complaining about "hallucination" is a pretty big "tell".
The conversation around LLMs is so polarized. Either they’re dismissed as entirely useless, or they’re framed as an imminent replacement for software developers altogether.
Hallucinations are worth talking about! Just yesterday, for example, Claude 4 Sonnet confidently told me Godbolt was wrong wrt how clang would compile something (it wasn’t). That doesn’t mean I didn’t benefit heavily from the session, just that it’s not a replacement for your own critical thinking.
Like any transformative tool, LLMs can offer a major productivity boost but only if the user can be realistic about the outcome. Hallucinations are real and a reason to be skeptical about what you get back; they don’t make LLMs useless.
To be clear, I’m not suggesting you specifically are blind to this fact. But sometimes it’s warranted to complain about hallucinations!
The comments section here is also quite a study. Many of the responses are so quick to dismiss the article with basically a "Well of course, this is all obvious, it's clearly due to [insert dubious reasons]." Thankfully not all of them: Some of the top comments show some genuine curiosity and deeper reflection, rather than just a pithy shoulder-shrugging dismissal. But it's definitely a pattern.
It's a little funny to read this and then immediately see comments in other posts lamenting how LLMs are always "confidently incorrect." I suppose LLMs really are just mimicking what they've been trained on.