I admittedly flip-flop on this quite a bit. I use LLMs in my work and for a bunch of pet projects I don’t really care much about.
For the pet projects, it’s not really a big deal. I’m the only user, and I as the user don’t care how it works under the hood.
Work is a different story. I have never felt more disconnected from my work than in the last 6-8 months. It takes a huge mental toll to read into our code and figure stuff out, almost like I just got hired and I’m working on new code for the first time. I’m asked to read “specs” written by LLMs that are unintelligible and I find my brain disconnecting immediately. Tickets, also written by LLMs pose the same problem.
I’m asking teammates questions because I want their answer and I’m told to just ask Claude.
None of this feels good, it may be where the industry is heading but as a professional it feels wrong.
A lot of the "none of this feels good" is that the models have been RL'd to produce longer and longer outputs. You don't start with a 1 paragraph explanation, being forced into a back-and-forth with the model so much anymore, unless you explicitly ask. Instead, it'll generate a 10 page report that you sorta-skim and the hooks of your attention skid off, and you prompt something vague like "make it shorter".
when llms write all the code, all we have left are the specs. ive found a lot of success applying software engineering discipline to spec writing and making it a human-only activity. from there hand the spec to LLM with a few guardrails around implemention and writing quality tests that prove the spec.
PRs review more quickly when you have human written spec diffs to orient you and guide (or skip) the code review.
a danger to avoid is writing specs which are basically just really bad code (prose is terrible code). you can work towards patterns to avoid this pitfall.
also, dont coflate spec with plan. spec is durable decisions. plan is execution to turn decisions into code. plan is disposable, dont check it in.
either way, find the human layer and dont use LLMs there. make all the important decisions in the spec, delegate the other decisions to the LLM.
i believe this is the new craft.
that said, i expect a small but significant exodus from software. a lotta folks came for the "programming" that is now being outsourced. were becoming so tired, just absolutely burning out.
i hope we will eventually realize its time to go, instead of giving up our souls for a fucking paycheck.
What I hate the most is people vomiting AI slop communicating with others. I have teammates who decided that an agent producing a wall of text summarizing changes in a pull request instead of a simple "LGTM" is a good idea. I have teammates who talk using AI almost exclusively (you can see a small tag "Sent using Claude" or "Sent using Cursor" under their messages)
Posing a well considered, concise and precise question only to get a wall of Claudeisms in response from a colleague is always demoralizing. Then you have to play the guessing game of 'is this detail in the response true, or did Claude make it up for them?'
> I have teammates who talk using Claude almost exclusively.
If somebody told that humans are a form of creature which can forego their most important and powerful capabilities just for convenience, I'd call the person saying that batshit insane, but here we are...
a teammate who exclusively writes slop PRs and copy/paste LLM responses to slack has demonstrated they can be replaced by an LLM integration to slack and github.
If your company cares enough/will allow it, make a written policy that authors are responsible for their code and they must still meet whatever quality bar existed at your company pre-LLMs. It's awkward, but I've seen friends whose companies have started to make it normal to say "I'm not reviewing this, it was clearly written by AI" the same way pre-LLMs you could also say "I'm not reviewing this because it's 5k LoC and needs to be split up into smaller reviews".
Then if people send you slop you can either point them to the guidelines or, if they don't stop, report them for violating workplace rules and make them suffer.
I'm feeling the same, personal projects have never felt better, but work has slowly felt more miserable.
Code reviews come back with a ream of issues that I just want to scream "I don't give a fuck".
Pre-LLM no-one would have cared that if another class was configured differently ( it isn't ) and called the new class in a different way ( it doesn't ), that you'd in some race-condition occasionally leak a resource, but only on shutdown when the program is about to terminate anyway.
That's the kind of issue that claude loves to find and write up in a code review with the exact same verbiage and write up style as if it had just found a naive SQL injection from string concatenation.
Prior to LLMs I doubt anyone would even have spotted the first issue, let alone bothered to write it up.
But instead it's another round of review, another round of trying to decipher the LLM's comments to discover whether it's a meaningful problem or a nothingburger.
I’m in both a lot of art based communities and programming based communities (like HN). The one blind spot many HN readers have is that the quality of something is one facet of an artwork.
Humans make minute decisions in their art whether to tell a story or represent something. When someone generates something, at some point you as the viewer/reader could assume that arbitrary choices were made in places. Those arbitrary decisions just aren’t interesting. I’ve quoted this before, but Kane Parsons highlights exactly this thing:
> If I see an element of the environment has been, you know they used generative fill or whatever to change something about the scene, even if--. It just shuts the part of my brain that wants to know more about that world and wants to look for details, 'cause I would assume if they're willing to make an arbitrary choice there, they can make an arbitrary choice with literally anything.
It’s just less interesting to our minds, I think because there’s less to ponder and explore. It can be a masterpiece in quality but when I know something in that piece was just made up by a machine, why would I care? I can’t wonder why the artist would include it. I can’t think about what it’s trying to tell me because it’s likely not telling me anything.
YES. Well said. This is why I want to know which text is from human thought labour. To read text or to consume outputs of human minds is to learn about those minds and how those minds might work.
There's a deep loss in having no mind to learn about in the subtext of the specific knowledge content written... I never knew this potential was there (for me) until it started to disappear as a domain of my interest and enthusiasm
I'm definitely a collectivist. Maybe this is related to that value system, or some other subtle thing that I didn't know varied in experience until a fracture and unbundling started. It feels a like how COVID introduced new fracture points in worldview and value systems that became less tenable to ignore, or at least harder to bridge and work across
I don't quite know why but something in the realisation that something is made with AI removes all the joy from reading/listening to/viewing/watching it. It was interesting reading your take on it, I'm not sure it's why I don't like it but I also don't have a better explanation.
It's not about skill either, I'm quite sure I'd react the same if AI writing surpassed Hemingway, it's still not art.
It also just occurred to me, inspired by your comment, that getting to know a mind is more than just satisfying your curiosity; personal experience confirms for me that as soon as I interacted with a brilliant mind, some of that shone back on me, and made me a tad more brilliant as well.
We always want to be among the people we imagine or visualise our future better selves to be. It's such an absolutely fundamental part of humanity (it is how children become adults, too)
The reason "AI artists" engage in a spiral of angrily and defensively explaining that their process is not just pushing a button, or trying to project AI-like attributes onto existing things like "covering someone else's song" (which is what the Fender CEO recently did!!) is that we all know that, deep down, using AI to do creative things people did is less in at least some way. It's avoiding some emotional labour.
People who know this less is a big part of what they do are the ones who need to reckon with it.
Everyone else is free to mentally downgrade it just like they also mentally downgrade excessive photoshop, overproduction, etc., and they will.
It took a generation for the world to find any kind of accommodation with what photoshop is capable of, what is appropriate and what is not, and there are aspects we are still justifiably angry about (fraud, deception, body image). AI artists seem in a hurry to have their own use normalised; they want people to hurry up and get over it and use it themselves so they don't feel bad.
Many photographers have developed rules for what they will and will not retouch, even a century before photoshop — for example, if the spot on someone's face won't be there in a fortnight, it maybe even shouldn't be in a photograph that is intended to be a long-lasting, somewhat authoritative representation of that person. But this took a long time to become any kind of cultural norm.
I think AI-only coders and CEOs are also in this "hurry up and get over it so I feel OK about my choices" phase, and I think we need to see this in a clear-minded way if we are to confront its impacts.
I think it’s a blind spot to a lot of people not just programmers. If you don’t have experience understanding and not just viewing you’ll almost always prefer the shinier object. I don’t really think this is particularly new even generative AI has just ramped up the speed that shiny objects can be produced. Most people aren't trained to do the understanding part so most media has always been about maximizing the shininess.
I’m sure general population won’t care but they’ve got me fucked up if they I’d ever use something like this. Glean my personal info and preferences the hard way! Through what insane videos I like on Instagram!
I had these features off on Windows where possible. However I can’t imagine using Windows these days. The only time I switch to Windows is to play Marathon/Destiny these days. Other than that, I have no reason to use it. I imagine so many others would have a similar experience.
I’m all for an UBI but a future where people have no purposeful jobs, no way to actually make a real difference to anything around them, is a bleak one.
I would love such a future. I would be perfectly content focusing on my hobbies and playing video games rest of my life. Perhaps the UBI would be sufficient even for a little bit of travel.
Even if in the unlikely case UBI gets implemented, there is no way payment will be equal. Why would the rich give up their positions of power? They’ll just get more security and bribe politicians, enough to build private armies and fortresses so the poor and starving can’t touch them, and have no choice but to live out their lives in squalor…we’re seemingly headed that way according to many even without pervasive autonomous machines taking away all work.
> The billionaires considered using special combination locks on the food supply that only they knew. Or making guards wear disciplinary collars of some kind in return for their survival. Or maybe building robots to serve as guards and workers – if that technology could be developed “in time”.
You can stream Big Picture with tools like Sunshine/Moonlight or their forks. Works amazingly well over a LAN if you have a server somewhere and stream to something connected to the TV. I have a headless server and an NVIDIA GPU so not sure what the sibling comment is referring to about compatibility.
For the pet projects, it’s not really a big deal. I’m the only user, and I as the user don’t care how it works under the hood.
Work is a different story. I have never felt more disconnected from my work than in the last 6-8 months. It takes a huge mental toll to read into our code and figure stuff out, almost like I just got hired and I’m working on new code for the first time. I’m asked to read “specs” written by LLMs that are unintelligible and I find my brain disconnecting immediately. Tickets, also written by LLMs pose the same problem.
I’m asking teammates questions because I want their answer and I’m told to just ask Claude.
None of this feels good, it may be where the industry is heading but as a professional it feels wrong.
reply