@ujjwalscript If you don’t review its outputs and don’t do any up front planning, sure.
But I know for a fact that it’s made me faster at shipping good code - and the tech debt is from product requirements changing, not the “vibe coding.”
English
Matthew Tanous
281 posts
Matthew Tanous
@MatthewTanous
Software engineer. Writer. Exploring the intersection of Scripture and economics - and what i means for the church to actually care for its people.
Chattanooga, TN Katılım Mart 2026
34 Takip Edilen1 Takipçiler
The "10x AI Developer" is a MASSIVE lie.
You are just a 1x Developer generating 10x the technical debt.
The entire tech industry is high on the illusion of "vibe coding" right now. The popular consensus is that because Claude and Devin can spin up a backend in 45 seconds, software is now infinitely cheaper to build.
Here is the provocative reality nobody is budgeting for: AI is about to make software engineering significantly MORE expensive.
Everyone is cheering for code generation, but completely ignoring the Verification Tax.
When an AI agent writes 5,000 lines of code, it is optimizing to pass the immediate test. It is not optimizing for human readability. It relies on brute-force loops, repetitive logic, and bizarre architectural shortcuts that just happen to compile.
Fast forward 12 months. Your business needs to pivot, or a core dependency breaks.
You are now staring at a 50,000-line black box that no human being actually wrote, understands, or can safely modify. You cannot simply "prompt" your way out of architectural collapse.
When the machine-generated spaghetti finally breaks, you won't be saved by a $20/month LLM subscription. You will have to hire a top-tier Principal Engineer at absolute premium rates just to untangle the mess your "autonomous swarm" created.
We are treating code generation as a pure productivity win, but code is a liability, not an asset.
Stop measuring how fast your team can generate syntax. Start measuring how quickly they can debug it.
English
@GalvinAlmanza It’s not that the AI replaces juniors, it’s that it means you don’t need as many. Like other technical advances, the belief is that these will save labor and increase productivity, so other jobs will be opened up.
The assumed scale is probably wrong, but it is happening.
English
I don't get how people are planning to sidestep the very basic problem that if you don't have junior hires right now, you won't have experienced people 5 or 10 years later.
CG@cgtwts
Anthropic CEO: “50% of all entry-level Lawyers, Consultants, and Finance Professionals will be completely wiped out within the next 1–5 years." grad students and junior hires are cooked.
English
@SpencerKlavan This is actively false, and stems primarily from methodological errors on the study designers that cause it to fundamentally misuse the tool for its claimed attempted purpose. “Edit this essay” with no context or guidance as to tone, voice, or intended audience is just dumb.
English
This is structurally necessary, totally predictable, and above all completely obvious to anyone with even a whiff of literary sensitivity
nxthompson@nxthompson
A new paper shows that AI does make your writing more homogenous—and it'll do the same even if you just ask it to edit your writing. Even weirder: it's not like it'll push you toward the voice of an average human, it'll push you toward a kind of AI-optimized voice that's subtly different from the way any people write.
English
5. No comparison to human editors who aren’t the original author - naturally the author already has contextual information about the essay. But another human may - and often does - suggest larger edits, which can also change the meaning if they misunderstand the original.
English
3. The system is configured to give the LLM free rein - no need for approval on the edits.
4. No tone or audience guidelines to the edit - another problem from vague instructions.
English
Obvious methodological errors here.
1. The prompts are shit - they don’t include any info about user intent, just “edit this” as one shots.
2. The essays are extremely short, and meaning is not clearly structured in them - so “altered meaning” is a vague statement.
nxthompson@nxthompson
A new paper shows that AI does make your writing more homogenous—and it'll do the same even if you just ask it to edit your writing. Even weirder: it's not like it'll push you toward the voice of an average human, it'll push you toward a kind of AI-optimized voice that's subtly different from the way any people write.
English
@KrakeJames @STOTTINMAD @Sargon_of_Akkad I tend to read multiple books at a time, keeping the context of each in my head as I go. You absolutely cannot do that with Malazan.
English
@STOTTINMAD @Sargon_of_Akkad Malazan needs a warning label attached to it that if you are starting and stopping it as an audiobook on your commute to work, it will make absolutely no sense and you need to read it in physical copy instead.
English
If you're going to create such a complex fantasy world, then you have to consider how your player characters will be situated in it. The DM here has created his hero to basically be an impoverished beggar, for some reason, rather than the son of a blacksmith or guild member--a man of means. Did the player just pop into existence with nothing? Does he have literally no relationships in this world, no connection to anyone or anything? No kidding your beggar PC can't afford a horse because you rigged the game by starting him with only 6cp and no help.
Moreover, this is a very silly and basic problem to overcome. Here is the solution (other than actually giving him a background in the world): a lord has heard of the player's quest, and approaches the PC to tell him that he will provide him with the necessary starting gear, on the provision that the PC brings back a particular amulet or something that can be found at his destination. The amulet has no real monetary value, but is a locket with a picture of his wife, who passed away sadly, and is a sentimental keepsake he had stolen.
Problem solved.
Orwell & Goode@OrwellNGoode
English
@KrakeJames @Sargon_of_Akkad Seems as basic as “protagonist already has a horse” or “quest giver knows the prophecy and supplies hero.” One might start to suspect this was Reddit clickbait.
English
@Sargon_of_Akkad This isn't for tabletop, this was posted as a 'problem' for a fiction novel. Which makes it all the more absurd this writer can't come up with a solution.
English
@nxthompson I shouldn’t believe my lying eyes then, when it clearly produces different results when prompted to use a different style of writing or for a different audience?
These studies are horribly flawed.
English
A new paper shows that AI does make your writing more homogenous—and it'll do the same even if you just ask it to edit your writing. Even weirder: it's not like it'll push you toward the voice of an average human, it'll push you toward a kind of AI-optimized voice that's subtly different from the way any people write.
English
@hai_torto @friedmann_20843 @uskglasses Seems like an overstatement to me. The initial entries in the genre (like Aura Battler Dunbine) were not conflict-avoidant, and not all of today’s more recent ones (Overlord, Tanya the Evil, Arifureta, etc.) have shifted that way.
English
@MatthewTanous @friedmann_20843 @uskglasses It’s still serving wish fulfillment in a different dressing though. Both refuse conflict as much as they can and see it as a chore.
English
one day you'll answer for the harm you've done
🌿💚Basil➡️HaloFest💚🌿@itsbasilbasil
There's a lot of modern fantasy that is obsessed with sanding down any emotional/sociopolitical tension by making everyone unproblematic and fluent in therapy-speak in ways that make zero sense for the setting or characters, while allowing the quirky fun violence to go unexamined
English
@friedmann_20843 @uskglasses Most isekai until recently were about power fantasies. Less found family, more “save the world for fame and girls.”
It’s been changing recently. But it maintains that core trope in a way the Western version doesn’t.
English
@uskglasses It's interesting to see that being blamed on critical role, as if most anime isekai aren't about cozy friendships.
It's simply a market.
GIF
English
@Darkzeid25 @uskglasses They’re frequently similar enough to compete for the same audience.
And can we go back to dragons that were fearsome and dangerous, not “bonded pets”. How To Train Your Dragon and Eragon did a number on those magnificent creatures.
English
@uskglasses I'll take a thousand cozy fantasies over another damn romantasy
English
@Blundarian @uskglasses The worst part is that it has essentially entirely replaced the possibility for literary fantasy that we used to see in novellas - small stakes, interesting stories involving magic and fantastical creatures. Now it’s all “I’m gonna open a store and have a queer romance!”
English
@uskglasses The cozy subgenre is one of the most repulsive things to come out of recent trends
English
@annielcrawford Sure, but I can be literate in the field and think Postman was wrong - that TV didn’t degenerate us because of something inherent to it, but we degenerated it because of something inherent to us as fallen humanity.
English
YES - if you are not literate in a field - media ecology / philosophy of tech - please don’t make yourself a public “thought leader” on things you don’t understand.
All the Christians, who are jumping on AI are basically repeating the same mistakes of all the televangelists of the 80s.
Shane Morris@GShaneMorris
There are just so many people in the AI convo who need to read Neil Postman. The constant assertions/assumptions that all technologies are neutral tools dependent on the user's morals are stunningly naive. You challenge that assumption & people keep repeating it like a catechism.
English
@sukh_saroy If it pushes towards agreeable text, why am I constantly editing out cases where it exaggerates a criticism I’ve made from “there is also this point contrary to the current view” into “and this is devastating for [the opposing perspective]!”?
English
🚨 BREAKING: You asked AI to improve your writing.
It changed what you were actually saying.
New research just proved it.
In a controlled study, heavy AI writing assistance led to a 70% increase in essays that gave no clear answer to the question being asked.
Not unclear writing. Neutral writing.
The kind that sounds polished but commits to nothing.
Here's what makes this worse:
Researchers took essays written in 2021 — before ChatGPT existed — and asked an LLM to revise them based on real expert feedback.
The instruction was simple: fix the grammar.
The model changed the meaning anyway.
Every time.
It can't help it. The training pushes toward inoffensive, agreeable, averaged-out text. That's not a bug they can patch. It's the objective function.
And then there's the peer review finding.
21% of reviews at a recent top AI conference were AI-generated.
Those reviews scored papers a full point higher on average.
They also placed significantly less weight on clarity and significance — the two things peer review is supposed to evaluate.
So we're not just talking about your email sounding a little corporate.
We're talking about AI quietly flattening scientific discourse.
Laundering opinions into non-answers.
Replacing your voice with the mean of everyone's voice.
The industry keeps asking: is AI-written content detectable?
Wrong question.
The right question is: what are we losing when a billion people let the same model edit their thinking?
English
@NidaKirmani “Displaced poor farmers” right into the middle class that was invented when industrialized farming meant that 90% of the population could do other things beyond scramble to feed themselves.
There would be virtually no artists to complain about AI without industrial automation.
English
A more apt analogy would be comparing those who refuse AI with those who resist corporate farming methods, which have decimated our food systems, displaced poor farmers, damaged our health, & destroyed our planet.
Stijn Noorman@stijnnoorman
People who refuse to use AI are like farmers who refused to use tractors.
English
@DanFriedman81 A decent novella trapped in quippy Millenial hell.
English
Writers like Dan Brown and Andy Weir, who write ugly and clumsy prose at the sentence level, are sometimes extremely engaging storytellers in ways that transcend their technical deficiencies.
“Project: Hail Mary” is a story about a lonely astronaut who undertakes a one-way mission into deep space to do a science thing that is the only hope to save the world, and when he gets there, he encounters an alien bug who is also a lonely astronaut on the same kind of mission, and they become best friends and work together to do their science thing and find a way to get back home.
It is utterly charming, and everybody knows it. All these dunks are meaningless.
Robert Rubsam@rob_rubsam
Driving me slightly insane that everyone has to pretend that Andy Weir is a real novelist and not a guy whose books are like 70% math equations
English
@jibbles_au @DanFriedman81 If you have to convey emotion or weight with seasoning verbiage, in most cases, you’ve written poor dialogue.
English
@DanFriedman81 Trying to explain this exact feeling to my friends after reading through Weir dialogue where six lines in a row are some form of "he said", "she said", and "I said." Common problem with scifi authors who often have great story concepts but average penmanship.
English
@LuinRomestamo @emperorthesteve @HariSel57511397 No, I don’t agree. I do not believe that anything has meaning just because it was worked on for a long time. The content of art is the meaning, not the time investment.
And mass production does not remove it, or the idea of publishing books or selling prints would be absurd.
English
@MatthewTanous @emperorthesteve @HariSel57511397 I actually does mean something, even if the book is bad. You're confusing meaning for quality. People can struggle for a long time to achieve something, and their achievement may not be of high quality. But it will always carry more meaning, more soul, than anything mass produced
English
Tell me you’ve never spent three nail-biting hours on a sentence that changes the whole meaning of the book without telling me.
Josh Daws@JoshDaws
Anti-AI writers treat their work like a horcrux, as if typing words manually embeds a piece of their soul into the draft, and AI use would corrupt the sacred vessel. This is not theology. It’s not even good metaphysics. It’s magical thinking dressed up as artistic integrity.
English
@Cyn_Cyb3r071Qu3 @DaveShapi It’s not lying. The two are generated statements with no real connection.
English
Why does Anthropic make their model deliberately lazy
English