@ludicity I would say the breakdown is more "somewhat useless vs an actual liability." It's not that thet aren't tons of smart capable people it's just that Jack Welch corporate engineering really benefits from somewhat useless but not dangerous. tons of exceptions but average is overall not that good

@ludicity pre-LLM? Under five percent were completely useless.

@ludicity usually they know and care about what they are paid to know and care about. Ah, wait. I only know underpaid devs, because I only know incompetent fucking managers whose shit these poor dev souls have to deal with.

@ludicity I've been in software engineering for 20 years and I haven't met any people like this yet, even by stretching the definition. Lots of completely unmotivated software engineers, lots with surprising and alarming holes in their knowledge, plenty of unreliable ones, but useless? Never. Not even the blatant copy-pasters were completely useless or completely ignorant

@ludicity "Software engineeering" knowledge or "software development" (aka "coding") knowledge? Lately (past 20 years), I've rarely run into any software engineering knowledge, but looking for books and courses, most seem focused on coding anymore.

Agile has killed most process movements, and the XP/test-driven crowd has been the loudest voice killing off analysis and design efforts. I can't blame them for running from the elaborative methods camp (Booch, Rumbaugh, etal)

@ludicity I wince imagining the responses you’ll get. But here’s mine.

Completely useless or lacking in basic knowledge? In an unexpected and undesirable way? Or in an expected and acceptable way? Every intern is this. Every junior engineer in their first role is this. Just like an apprentice in any trade. Every experienced developer who moves from a domain they know well into a domain they don’t know at all is this.

I assume there’s an implicit “when they’re not supposed to be” in the question. Or “and they’re unwilling or unable to improve enough” in the question. I’m just not sure the goal or the point of the question. I’ve run into lots of these folks over the years, but mostly in places where they’re acceptable and expected. If all is well, we bring on people who have good potential when they’re hired, but no skills, yet. And we remedy that through training and experience.

Like, someone comes in knowing language syntax and coding basics, but they have only worked in toy projects, never million-line systems that are 15 years old and the product of multiple acquisitions. They’re useless on day one in that kind of environment. But they learn (if all is well).

@ludicity In a global way, yeah, rarely, except for a few cases that didn't last long.

But what i've noticed now is engineers that are very competent in their specific part of the work, but will now mess with things they have very little knowledge of, without taking time to understand such thing, and applying fixes/making changes to those things.

Sometimes without informing those in charge of those changes because "It was just a quick fix."

@ludicity the software engineer who doesn't know anything has always been part of the industry, and will be so long as it pays well enough to attract people with no particular talent or passion for the work. Yesterday's "lmao I don't know anything, I just copy from stackoverflow" is today's LLM dipshit.

The guy who sat next to me left the company before a year was up. He couldn't do anything without the LLM. Couldn't write a document, couldn't program, couldn't search a text file for lines matching a string. One time he said he realized he wasn't learning anything about our code base by having the LLM do his work for him, so he started asking the LLM to summarize the code. Data point of one, but oh my god

@ludicity I don't think I've ever met a software engineer who was completely useless. In my experience, the bigger problem is people who overestimate their abilities. But I think that's a phase many of us have gone through at some point in our careers.

@ludicity I deal a lot with the new species of “cosplay developer“ since the advent of LLMs as an architect.

Until last Fall these PMs-turned-expert let ChatGPT write meaningless 1000 word Jira tickets full of erudite technical prose of nothingness to feel important. You rewrite from scratch, only twice the effort. Fine.

Now they‘ve learned to install Claude Code and Openclaw and feel like their opinion in development is useful. So infuriating.

@ludicity I'm lucky to write code for embedded devices and systems with a team that has been doing it since before llms were a thing. When/if we bring in new devs it will be interesting to see how much they rely on llms

@ludicity Rarely, indeed.

In "the old days", people were hand-picked (hire for attitude, train for skill) and very few fell behind over time.

With LLMs around, we decided to introduce an #AIBadge, a qualification you need internally to use AI for manipulating product code. Generate all the unit tests and dev docs you like, but you have to prove you can do an outstanding dev job without AI before you get to use it.

That way, we are still training up our new recruits to a level where it's reasonably safe to use AI, and we don't lose our source of qualified engineers.

Luckily, our business model does not force us to go all-in with AI. Plus, freeing up time from the most experienced engineers gives us the biggest edge, speed-wise.

@ludicity Not as rare as you might hope

Before LLMs - Corporate IT hides a lot of frankly terrible developers. I've had people who have either viewed it as a way out of service desk hell, or as some kind of stepping stone into Project Management (WTF?) and it's easy to use the corp machine to hide

Neither me nor my current team use LLMs (and my current - small - team AREN'T USELESS) but my gut feel is that, of the terrible devs I've known, they would find it easier to hide behind LLM code

@ludicity yes, a mixture between fresh from college interns and junior security consultants from a big consulting firm, who were sent to us to train them up to blue team security consultants at banks

That they were juniors were, obviously, the reason that they were useless

I teach college electrical/computer engineering students programming, and they've completely lost problem-solving skills. like, actually! these people are going through my class to learn C, assembler, and how a computer's CPU works along with its peripherals, and these students are BEYOND cooked.

they, quite literally, cannot troubleshoot a simple C program without the aid of ChatGPT. they cannot read code. they cannot go through the motions of problem solving! I throw code up on the whiteboard, and two students whip out their phone to ask ChatGPT to transcribe the code (which, of course, it fails miserably at because it did not understand two-column formatting). "oh, help, my motor controller circuit doesn't work!" (the solutions are literally in the lab manual) "let me take a picture of a breadboard with a bunch of chips, wires, a microcontroller, power transistors, resistors, and capacitors and ask ChatGPT how to fix this!"

oh, the prerequisite to my class is a course on ARM assembler programming. I loved that class, because we used the oldskool ARM assembler as opposed to the GNU assembler, but also because the problems were fun to solve -- we had to write all manner of assembly programs, and there was one due every Tuesday and Thursday. unfortunately, the students in that class taking it in the post-LLM generation have done essentially none of the work themselves; the test averages are comparatively low, and cheating is RAMPANT.

in my class, I routinely get submissions for assignments that contain either total AI-generated slop, accidentally contain "as a large language model", use oscilloscope screenshot pictures I saw in spring 2024, or do not mirror the voice of the person who supposedly wrote the assignment.

so, how often do I encounter people that seem completely useless or lack basic knowledge? every day!

these people are seniors. they graduate 8 days after they give their final presentations.

@ludicity

all. the.time. Both before and after LLMs.

Devs with minimal networking knowledge.

Web-devs with zero HTTP knowledge.

C coders writing buffer-overflows and failing proper malloc/free pairing

SQL injection vulnerabilities out the wazoo.

Interviewed folks who couldn't write a basic fizzbuzz loop (or similar Coding 101 style exercise) in languages they claimed to have expertise.

I lived through the years of "I can use Frontpage and FTP files up to a server, I must be a web-developer!" with no regard to semantic markup, accessibility, security, usability, etc.

@ludicity I'm seeing "senior" software developers who think feeding the entire codebase into an LLM is the way to get to grips with the code, and don't notice when the unit tests only check the math library works not any of the things our code actually does.

I'm glad to be near retirement as I really can't take this for much longer

@ludicity My worries about LLMs in SW eng is not with software engineers, who usually have the basics and expeience required to avoid the biggest LLM issues. No, my worries is with *NON*-SW-engineers who believe themselves capable of doing SW dev with the "help" of LLMs, but who lack even the fundamentals about SW dev and thus commit (pun intended) horrible things (or would commit if only they knew about source code management).

@ludicity Before LLMs, when I’ve been involved in interviewing, I’ve seen lots. We (almost) always washed them out at interview, but bear in mind these are not people who are fresh out of university or college. They sometimes have a decade of “experience”.

I think a combination of remote interviews and LLMs are allowing more of them to appear competent and sneak into organisations that would have previously rejected them.

@ludicity working in a company with ~400 engineers. Colleagues generally good, but I also interview people. The quality of candidates has dropped significantly. I can't say for sure if this is due to other factors (are we offering sub-market rate, we switched from mainly Scala to Typescript...) but candidates are not allowed to use llm in interviews and some seem completely lost.

@ludicity I've done technical interviews of 100+ candidates for software engineering roles over the years. Around 90% of them didn't pass muster. And that's after a CV screening by HR (for whatever that's worth) and a preliminary interview with a teamlead.

It is RIDICULOUS how many career programmers are, in fact, not able to program anything in their language of choice. I'm not talking about having to google how to open a socket, but about how to write a god damned loop ... Awful incompetence.

@ludicity Rarely proportionately, frequently absolutely.

For the benefit of people (like me) who never heard of Ed Zitron:

<https://www.wheresyoured.at/about/>

<https://redd.it/1rak1tp> includes a link to his post on Bluesky.

#LLM

@ludicity 10 years in data engineering. "Completely useless" is something I've only seen a few times, more often in non-technical managers which wasn't the question.

I do often see engineers who don't understood best practices or good architecture. Or don't understand the frameworks they are using. Or frankly just don't try.

The LLM spell mostly affects the beginner or mediocre engineer. Senior engineers find them mostly frustrating but occasionally useful.

@ludicity i don't quite know how to tell ed that, like basically every other field of endeavor, software is permeable to people who have no useful idea what they're doing.

(or, i guess, that some of the people who lack basic knowledge and have no ability to contribute will probably stay that way forever but that many others eventually figure things out and become pretty effective.)

@ludicity whats your profile on bluesky?

@ludicity it comes in waves. Working in SF 1998-2001 anybody that could hack out HTML was a “developer”, then 2010-ish the proliferation of “code camps” where everybody was a “full-stack developer” because they could have a LAMP stack hack out HTML…

@ludicity A number of times, but it mostly seemed to depend on the shop, its culture, and how they hired. The worst have been the mid-size agencies, which tend to optimize for minimizing inputs while maximizing billable hours, resulting in software that *mostly* works, with source code that looks like it was written by a dog. Such shops absorb large quantities of useless engineers from the market - mostly those that got into it for the money.

@ludicity fortunately none so far. It's still a relatively short career though

@ludicity pretty often

@ludicity I've worked with people of a wide range of abilities, but not sure I can describe any of my former or present colleagues as completely useless I've

@ludicity

We must welcome folks with no experience, and not deride them as being “useless”.

Lack of compassion and human engagement, and the capitalists dream of the 10x hero programmer got us into this mess.

It’s your job to develop your team. Train them. Believe in them. Support them.

It’s not a pissing contest.

@ludicity I'd say "occasionally", and the rate at which it occurs is notably distinct between different kinds of workplaces. The more enterprisey, the more likely.

Also, interview candidates seem disproportionately more likely to be terrible, but that at least makes sense--the less competent folks are more likely to be interviewing, right?

I wouldn't say I've noticed much different pre- and post-LLMs, but I HAVE noticed that post-LLM, you've got a lot of people uncritically drinking the Kool-Aid.

@ludicity bear in mind a lot of tech companies are casually relocating PM talent into SWE. Happened at the huge huge tech co I work for. All of a sudden I'm expected to start building things. Never been a coder. But it's fine because AI will help, right? Everyone needs to pay their mortgage and the job market is hard right now. So here we are.

@ludicity Completely useless? Not a lot, either pre/post, but I mostly worked in safety related industries. There were/are quite a few, which seemed to lack any kind of foresight or initiative.
AI use is picking up steam though, and I do see people outsourcing their thinking, sometimes live during a meeting (and getting obsolete data because of that).

People that are unable to see risks in software will use GenAI to generate a lot of security issues. Not looking forward to cleaning those up.

@ludicity I can't think of any. A few who weren't very good; a couple who spammed the codebase with buggy code that took considerable effort to clean up; and one or two who were good at day-to-day programming but didn't know much computer science, so sometimes wrote code with terrible big-O. But nobody who was "completely lacking in basic knowledge". I've interviewed a few candidates who appeared not to be able to code, but maybe they just couldn't code *in interviews*?

@ludicity So it's a difficult question, because on the whole programmers can be judgemental pricks.

The important metric is less about being "useless", but more "in the long run, do they provide value greater than their salary". In my experience, there's about 20% of the workforce who fail to clear that bar... but I also think it is actually hard to improve on that.

Given human nature, the ability of people to bullshit, and the general failings of HR depts, there will always be a dysfunctional part of the workforce.

@ludicity i'm a bit worried about confirmation bias here, though of course incompetence has existed and will continue to exist. the difference between a competent and incompetent engineer isn't decided by the tools that they have access to but the time they choose / are afforded to develop competency and how well they have learned-to-learn.

that said, while there isn't a quantitative difference in incompetence engineers, there is a qualitative difference in incompetent engineering. expensive AI licenses move wealth from labour to capital and give management hacks a license to demand specific things from engineers at a specific rate. some of the heaviest AI users ive seen are the junior enggs and interns, and while they werent able to answer questions about what they wrote pre-LLMs either, now it's buried in an amount of noise and unaccountability that makes it hard to catch these pitfalls during code reviews.

LLMs dont make people incompetent the moment you touch them. they change the amount of code, plausibly functional code mind you, that you can create in a given amount of time. this reduces the amount of time seniors can spend in design, reviewing, and talent building, and hinders the processes that (sometimes) build competence out of incompetence. i'm not a full-time-hater of LLMs, but i do worry about the real damage they do to enterprise engineering processes moreso than the engineers themselves.

@ludicity all I am learning here is I better brush up on FizzBuzz if I want to do an interview again.

Will send you a better answer privately, maybe.

@ludicity I worked for most of my career at a place that had a very good interview process, and pretty much everyone was competent on all the right axes. But at one point we started using some contractors from an agency and quickly realised we had to do our own screening of them. I usually asked them to code FizzBuzz in their choice of language and explain what they were doing as they did it. 20% couldn't do it at all. 30% struggled to explain their reasoning or listen to hints.

@ludicity oh boy. Pre, regulary, absolutely. LLM do not seem to have made that much of an inroads yet into our field, except Juniors getting led astray and eventually coming back very confused.

@ludicity I don't think it's happened in my professional life. At each company I've worked at there are some programmers who seem to be a bit behind the curve, and occasionally a few who don't do very good work, but nobody I would consider completely useless.

@ludicity asking this question speaks inexperience loudly. Incompetence is widespread in all areas of life. Even before LLMs. Especially in enterprise.

@ludicity A hundful, maybe two or three over the span of 10 years.

@ludicity It wasn't great before, but I've only seen one very specific slice of the tech world. I've encountered developers using technology they didn't understand. I've received too many *screenshots of stack traces* from developers on other teams, and they expected me to solve their problem for them. (Stack traces will, conveniently, show you exactly where the error is. And also it's your code.) I don't have super powers, I just know how to read and... program computers.

@ludicity
pre LLM: rarely in ooen source, often in corporate.

Now: likely in open source, mainly as security reporters who play copy&paste monkey with our project and their LLM. Cant say anything about corporate as I no longer experience that (thank the heavens).

@ludicity I've mostly met great people, before and after. maybe I'm lucky

@ludicity Depends. Rarely professionally, but I did most of my hiring for most of my life and I’ve seen things you people wouldn’t believe during the interviews.

The worst people were exactly like LLM - stupid, loud and unable to admit they are wrong.

@ludicity more frequently now, and specifically with software engineers who already had a lot of experience beforehand, but seem to be losing all their knowledge and best practices and making far worse choices when it comes to their code nowadays.

@[email protected] online? incredible common, IRL? rarely have i met someone who doesn't have a decent grasp of the tech they're using

@ludicity

See this post about -1x programmers:
https://infosec.exchange/@david_chisnall/116085039513622322

@ludicity maybe I moved in rarefied circles, but the big divide I've encountered is software engineers who understood things in general vs application "developers" who only knew one stack or app, and would frustratingly be convinced the sun shone from Redmond or wherever.

The advent of LLMs has disrupted this a bit, as they're more likely to answer in general terms, or contextualise the offering from the blessed stack as a poor imitation of a far superior open source equivalent

@ludicity 80% (and I’m likely being generous) are absolutely useless, lacking in necessary skills and basic knowledge.

@ludicity of the LLM developers I've run into, (which hasn't been that many since I've been avoiding them) they haven't been lacking in knowledge exactly, it seems to be something deeper.

Like, they mostly seem to know their stuff, but aren't used to thinking about how their code works or the structure of it, or how people are supposed to understand it, it just seems to get layered and layered, new bits added, and nothing really taken away.

Debugging and coherence seems to become an issue as they progress, but I think most teams will still have a couple of people that are capable of debugging (that's been me on previous teams), but if you haven't got that skill available, I suspect it'll become an issue in a lot of areas.

@ludicity For the record, I work at a software company that employs ~10k developers.

Before LLMs, I'd encounter such engineers a couple of times a month, but I interact with a lot of engineers, specifically the ones that need help or are new at the company or industry at large, so it's a selected sample. Even the most inexperienced ones are willing and able to learn with some guidance.

After LLMs, there's been a significant uptick, and these new ones are grossly incompetent, incurious, impatient, and behave like addicts if their supply of tokens is at all interrupted. If they run out of prompt credits, its an emergency because they claim they can't do any work at all. They can't even explain the architecture of what they are making anymore, and can't even file tickets or send emails without an LLM writing it for them, and they certainly lack in any kind of reading comprehension.

It's bleak and depressing, and makes me want to quit the industry altogether.

@ludicity About 90% of the time, so goes Sturgeon's Law.

@ludicity Sadly, the answer is not “rarely”. In the past few years, I’ve been fortunate enough to work in a fairly stable group, so can’t really differentiate between before and after LLMs. But regardless, there is a reasonable fraction of NNPPs — net negative productivity programmers.

@ludicity I founded an entire subreddit on this topic which still considers me to be an absent god.

Truly. But its name cannot be spoken.

@ludicity Like… basically everyone I’ve interacted with in my career except a notable handful?