sigotirandolas

Perhaps that's true of the top 5% engineers. But the question is if the top 40% of engineers can replace the bottom 60%.

The LLMs don't need to be perfect, they just need to be good enough so that the cost of fixing their code is lower than the communication overhead and the 'lost in translation' overhead from delegating tasks to mediocre engineers.

I don't look at whether the text is written by an LLM but at whether it has substance and whether the writer understands what they are doing and is respecting my time.

If the text is full of punchy three word phrases or nonsense GenAI images then that's an obvious sign. But so is if the other person has some revolutionary project with great results but they can't really explain why their solution works where presumably many failed in the past (or it's a word salad, or some lengthy writing that doesn't show any signs of getting you to an "aha, that's some great insight" moment).

A good sign is also if the author had something interesting going before 2022, and they didn't fall into the earliest low quality LLM waves. Unfortunately some genuinely talented people have started using LLMs to turbocharge their output while leaving some quality on the table nowadays, so I don't really know. I'm becoming a lot more sceptical of the Internet, to be honest.

Sorry, I think you're right that I misinterpreted your comment. I still had in mind OP's example (BDD, mutational testing, all that jazz). I apologize!

Reading your comment, it looks like you work for a pretty nice company that takes those things seriously. I envy you!

My concern was that for companies unlike yours that don't have well established engineering practices, it _feels_ that with AI you can go much faster and in fact it's a great excuse to dismantle any remaining practices. But, in reality they either doing busywork or building the wrong thing. My guess is that those are going to learn that this is a bad idea in the future, when they already have a mess to deal with.

To put what I mean into perspective... if you browse OP's profile you can find absolutely gigantic PRs like https://github.com/leynos/weaver/pull/76. I can not review any PR like that in good faith, period.

On the one side I reject that product and engineering concerns are separated: Sometimes you want to avoid a feature due to the way it will limit you in the future, even if the AI can churn it in 2 minutes today.

On the other side perhaps your company, like most, does not know how to measure overengineering, cognitive complexity, lack of understanding, balancing speed/quality, morale, etc. but they surely suffer the effects of it.

I suspect that unless we get fully automated engineering / AGI soon, companies that value engineers with good taste will thrive, while those that double down into "ticket factory" mode will stagnate.

To be devil's advocate:

Many of those tools are overpowered unless you have a very complex project that many people depend on.

The AI tools will catch the most obvious issues, but will not help you with the most important aspects (e.g. whether you project is useful, or the UX is good).

In fact, having this complexity from the start may kneecap you (the "code is a liability" cliché).

You may be "shipping a lot of PRs" and "implementing solid engineering practices", but how do you know if that is getting closer to what you value?

How do you know that this is not actually slowing your down?

Luckily those I work with are smart enough that I've not seen a PR thrown away yet, but sometimes I'm approving with more "meh, it's fine I guess" than "yeah, that makes sense".

The most annoying thing is that even after cleaning up all the nonsense, the tests still contain all sort of fanfare and it’s essentially impossible to get the submitter to trim them because it’s death by a thousand cuts (and you better not say "do it as if you didn’t use AI" in the current climate..)

N~10^(6.5) aliens.

I hope this too but it's not a given, IMO. Previously people without technical chops failed quickly by being unable to deliver working code, now they can deliver mediocre code with the damage only becoming clear years later. It breaks the "can deliver code --> good technical ability" proxy and even after the initial damage wave, it's unclear if we will find a better proxy.

This looks like LLM slop (not only the writing, but the analysis itself).

To start, it is based on a single machine check. It has little context, but if this was a common problem, I'd expect more data points.

The MC happened 8 hours after the freeze. It's not unusual that a hardware/kernel failure cascades to multiple subsystems so I'd be sceptical that the MC has a direct relationship to the root cause of the freeze.

(Part 13) Why does it quote an Engineering Change Notice rather than a consolidated spec?

(Part 11) LaneErrStatus=0xFFFFFFFF is 32 bits. As far as I know, PCIe x32 is very rare.

(Part 8.4) How is it surprising MMIO isn't included in memory dumps?

(Part 4) How is the definition of a MCE relevant here?

The security model is that applications run in a sandbox (e.g. Flatpak, snap) and only get D-Bus, Wayland, etc. access via restricted means (e.g. xdg-dbus-proxy).

The "friction" is that Wayland developers don't want a sandboxed application with access to the Wayland socket to pwn your machine.

Trying to isolate applications within the same UNIX user is essentially unfixable since there's ptrace, LD_PRELOAD, /proc/$pid, .bashrc drop-ins, etc.

The author must know about all of this, as it's mentioned in the LD_PRELOAD note in the end. In my view, the model he proposes is security by obscurity (putting hurdles on top of a fundamentally insecure system).

In the short term, definitely.

In the long term, once the damage from vibecoding is better understood (for customer impact and team morale), there's an incentive to push them out, both from the leadership and the individuals side.

There's definitely some that hold CQRS, DDD, TDD, ... as _the_ way to design software and over-engineer around it, so I can understand some pushback.

Knowing those patterns is very helpful as a way to think about design problems, as long as you have the common sense to realize applying the pattern "by the book" is often overkill and you can just take some ideas out of it.

That article conflates as "Pure engineering" both reducing a software system to a small set of cohesive concepts, and architecture astronauts, when those are polar opposites.

When all leadership is asking is "what is the short term business value?", it's pointless to make that case. It's much easier to measure "yet another feature" than "fix the root causes of what makes our product subpar and slows us down". Not only that, but an incompetent engineer's "tech debt grooming" may make things worse.

I think that this may eventually become better now that there isn't so much dumb money around (no ZIRP) and with AI assistants taking on some low-effort work (enabling companies to lay off incompetent engineers). But it will take many years for companies to adapt and the transition won't be pretty.

I assume it can be different for everyone. This post resonates with me, but my social anxiety mixes being sensitive to negative feedback and low self-esteem.

So, you want to avoid both being disliked, but also being liked - because this puts you in novel situations you fear lead to an even bigger failure down the road.

> Maybe things are this equal and fair on the senior, high-paying part of the spectrum

I don't think the fundamental dynamics change by seniority, just that after some level there may simply be a smaller pool.

From the interviewers perspective, it makes sense to reject a candidate if they see any possibility it could be a flop. A bad hire is going to frustrate the team and look bad to the company, missing the best candidate is just going to result in hiring their next best pick.

> As someone just starting out, the general feeling among my peers is that I must bend to the interviewer's whims, any resistance or pushback will get you rejected.

I guess this is very context dependent but I can also see "bending to the interviewer's whims" backfiring if they see you're just trying to flatter them. I could see some interviewers valuing that you can explain your point if it's framed in a way that shows you are both observant and easy to work with. If it's framed as a more aggressive kind of pushback, yes that's going to get you rejected.

But yeah, I can also see that if you're willing to take any offer at any company as a junior just to get your feet into the industry most interviewers may not be specially smart and resisting is likely to go wrong.

> - How do the end user protect themselves at this point? Especially the average user?

- Install as little software as possible, use websites if possible.

- Keep important stuff (especially cryptocurrency) on a separate device.

- If you are working on a project that pulls 100s of dependencies from a package registry, put that project on a VM or container.

Not the parent, but the default `npm install` / `yarn install` builds will ignore the lock file unless everything can be satisfied, if you want the lock file to be respected you must use `npm ci` / `yarn install --frozen-lockfile`.

In my experience, it's common for CI pipelines to be misconfigured in this way, and for Node developers to misunderstand what the lock file is for.

> Measuring and feeling are not mutually exclusive.

They are not mutually exclusive, but they compete to a degree. If someone's time is mostly spent on what can be measured, they can't spend time on "common sense" or investigative work that is less easily tracked. At the end of they day, trying to measure everything makes as much sense as trying to document every line of code. (Most of this, naturally, also applies the other way around).

> This is just the frame that the author is trying to prop up in order to sell us their shallow, meaningless piece.

> I see this pattern a lot in “influencer” content that people sometimes share with me

I think a lot of the shallowness is from blogs or HN being a public, persistent, broadcast written media. In a face to face conversation, you can generally follow up and share more specifics and nuance without fear of getting a bad reputation.

If anything I think the bias is the other way around, on the Internet whatever you write can get cherry-picked and framed to make you appear terrible, in person it's much easier to get a fair sample.

For as much as the author may get roasted for stating the obvious, I've often seen this "measure everything" mindset, coming from those you'd think should know better than that.

I've even seen this stupidity in myself sometimes. In a way it's funny how you can get so lost on the numbers that you forget about the thing.

I'd say the opposite, LLMs are a know-it-nothing machine to perfectly suit know-it-alls. Unlike a human, it isn't that hard to get the machine to say what you want, and then generate enough crap to 'defeat' any human challenger.

My only answer to this is, the ones at the top up to the CEO must be mindful enough to realize this, smart enough to figure out a solution, and brave enough to act on it.

Otherwise, it's a matter of time until the house of cards falls down and the company stagnates (sadly, the timescales are less of a house of cards, and more like a coal mine fire).

> What i got back was the typical overly verbose and articulate review from chatgpt or some other llm. I thought it was pretty funny that they thought it would work let alone be acceptable to do that.

Did that end up working for you?

I had this same experience recently, and it floored my expectations for that dev, it just felt so wrong.

I made it abundantly clear that it was substandard work with comically wrong content and phrasings, hoping that he would understand that I trust _him_ to do the work, but I still later saw signs of it all over again.

I wish there was something other than "move on". I'm just lost, and scarred.

What makes me not want to use Hyprland is that the code has all kind of "YOLO" tells, the kind of ones that make you wonder if something is going to happen some say... for example:

- https://github.com/hyprwm/Hyprland/blob/00da4450db9bab1abfda...

- https://github.com/hyprwm/Hyprland/blob/00da4450db9bab1abfda...

- https://github.com/hyprwm/Hyprland/blob/00da4450db9bab1abfda...

A lot of it is being able to make reasonable decisions under novel and incomplete information and being able to reflect and refine on their outcome.

LLMs's huge knowledge base covers for their incapacity to reason under incomplete information, but when you find a gap in their knowledge, they are terrible at recovering from it.

To be devil's advocate, this is the kind of all-talk argument the parent was referring to. Once the paid option is available, people will demand it to be [cheaper / better / someone else] and still not pay.

While I don't love my money going to Google, I find YouTube's overall quality astronomically higher than Instagram/Twitter/TikTok/etc. and the amount of censorship/"moderation"/controversy has been relatively limited. When I find something I really want to keep I have always been able to download it without much trouble.

My thought is that we humans are bad (by computer standards) at arithmetic and memorization because those are not evolutionarily useful on their own.

On the other hand general problem solving is, and so far any attempt to replicate it using computer algorithms has more or less failed. So it must be more complex than just some simple heuristics.

Perhaps the answer is just "more compute" but the argument that "because LLMs somewhat resemble human reasoning, we must be really close!" (instead of 25+ years away) seems wishful thinking, when:

(1) LLMs leverage a much bigger knowledge base than any human can memorize, yet

(2) LLMs fail spectacularly at certain problems and behaviours humans find easy

I think the more realistic argument is that the model can generalize, but only by learning shortcuts (e.g. how to pattern match a problem to a likely answer) and simple algorithms (e.g. how to propagate carries in a multiplication). And this only appears intelligent because this pattern matching is really good and backed by a huge amount of compressed/memorized answers.

The AI pessimist's argument is that there's a huge gap between the compute required for this pattern matching, and the compute required for human level reasoning, so AGI isn't coming anytime soon.

I'm not convinced by this argument. You can fit a bunch of books covering up to MSc level maths on less than 100MB. After that point, more books will mostly be redundant information so it doesn't need much more space for maths beyond that.

Similarly TBs of Twitter/Reddit/HN add near zero new information per comment.

If anything you can fit an enormous amount of information in 1MB - we just don't need to do it because storage is cheap.

You'd be surprised. Off the top of my head:

Many are conditioned to see `x` as a fixed value for an equation (as in "find x such that 4x=6") rather than something that takes different values over time.

Similarly `y = 2 * x` can be interpreted as saying that from now on `y` will equal `2 * x`, as if it were a lambda expression.

Then later you have to explain that you can actually make `y` be a reference to `x` so that when `x` changes, you also see the change through `y`.

It's also easy to imagine the variable as the literal symbol `x`, rather than being tied to a scope, with different scopes having different values of `x`.