#ai generated code

[aiweirdness]

“Slopsquatting” in a nutshell:

1. LLM-generated code tries to run code from online software packages. Which is normal, that’s how you get math packages and stuff but

2. The packages don’t exist. Which would normally cause an error but

3. Nefarious people have made malware under the package names that LLMs make up most often. So

4. Now the LLM code points to malware.

https://www.theregister.com/2025/04/12/ai_code_suggestions_sabotage_supply_chain/

[aeolianblues]

I don't like that the dev community picks on people who are most fluent in Python, when the ChatGPT-using "vibe coders" are right there. At least Python babies are coding. Bully the non-coders instead.

[lessons-from-moths]

i am having the most delicious (vile) thoughts about Ford accidentally killing Stan when they reunite in the 80s, tries to bring him back to life and fails, and decides to try the only resurrection that he can reasonably perform.

So Stan wakes up, incredibly fuzzy on what happened. He remembers getting Ford's message, remembers parking his car and being cold and hungry and sick, but nothing else. And now he's staring at a ceiling, feeling...nothing. He feels nothing.

He bolts upright, gasping, trying to feel anything at all, like air in his lungs or a bed beneath his fingers, but there's nothing. He tries to scream. He doesn't have the vocal chords to do it with. He wildly looks around to see Ford frowning at him, standing by the table that he's currently lying on. He begins to speak into a small recorder that he holds in his palm.

"Attempt 14. Stan appears distressed. There seems to be an issue with the cognitive dissonance of biological functions. On the plus side, he's acting like he's alive. We'll need to tweak his cognitive functions, maybe bypass the physiological needs in his thinking. Commencing shutdown."

The world goes dark.

Ford returns to the computer that holds a program that McGucket helped him make, adding a line of code to the program, chewing on his lip in concentration. Once he perfects it, Stan won't even know that he's a machine. He can bypass those thoughts with a few changes to Stan's code.

As far as the vessel goes, McGucket is great at this sort of thing. He can manufacture a human body, or at least something that works like one. Any other blemishes can be written away in Stan's programming so that he doesn't realize that he isn't quite human.

Ford can do it. He can bring Stan back to life in every way that matters. He might be responsible for killing his brother, but he won't be the cause of Stan's death. Stan will live on, probably forever, exactly as Ford remembers him. He'll do it right this time. He'll be a better brother, now that he has this second chance.

[matthewsmania]

OH MY GODDDDD I CANT BREATHE

[starssoblue]

"the reason adrien is just instantly good at everything he tries is because he is programmed to be that way as a senti" aside from the fact that i don't think that's how it works (and also while he was decent at everything he tried with marinette he wasn't instantly good at all of them, and what marinette actually said to him was that he could improve in anything with practice but it was a great first attempt) did we all collectively forget about how adrien actually canonically isn't the best singer?

[dukemarjorie]

AI Concept Art of my Fan Story: STAR WARS SEEDS OF HOPE: What Happened on Cato Neimoidia. Introducing my original character First 2 chapters are up. Check out the story here.

[lemedstudent2021]

where is the gtfo button on this omg

STOP IT GO AWAY *bug spray*

guys help ive removed the gemini shortcut (thank u sir) but im still getting this when i look stuff up. help :')

[findingambrose49]

CHAVFORMATION is waiting

[ottopilot-sfw]

Vibe coding something. This shit is supposed to take my job? All y'all are fucked. We are all fucked.

[mintjuliee]

Support Artists Not AI bracelet now in my shop!

Are you tired of AI slop invading your favorite creative spaces? Do you yearn for the days when art was considered to be a byproduct of the human creative process instead of code? Me too, man, me too.

[jcmarchi]

Can Developers Embrace “Vibe Coding” Without Enterprise Embracing AI Technical Debt?

New Post has been published on https://thedigitalinsider.com/can-developers-embrace-vibe-coding-without-enterprise-embracing-ai-technical-debt/

Can Developers Embrace “Vibe Coding” Without Enterprise Embracing AI Technical Debt?

When OpenAI co-founder Andrej Karpathy coined the term “vibe coding” last week, he captured an inflection point: developers are increasingly entrusting generative AI to draft code while they focus on high-level guidance and “barely even touch the keyboard.”

Foundational LLM platforms – GitHub Copilot, DeepSeek, OpenAI – are reshaping software development, with Cursor recently becoming the fastest-growing company ever to get from $1M in annual recurring revenue to $100M (in just under a year). But this velocity comes at a cost.

Technical debt, already estimated to cost businesses upwards of $1.5 trillion annually in operational and security inefficiencies, is nothing new. But now enterprises face an emerging, and I believe even greater, challenge: AI technical debt—a silent crisis fueled by inefficient, incorrect and potentially insecure AI-generated code.

The Human Bottleneck Has Shifted From Coding to Codebase Review

A 2024 GitHub survey found that nearly all enterprise developers (97%) are using Generative AI coding tools, but only 38% of US developers said their organization actively encourage Gen AI use.

Developers love using LLM models to generate code to submit more, faster and the enterprise is geared to accelerate innovation. However – manual reviews and legacy tools can’t adapt or scale to optimize and validate millions of lines of AI-generated code daily.

With these market forces applied, traditional governance and oversight can break, and when it breaks, under-validated code seeps into the enterprise stack.

The rise of developers “vibe coding” risks supercharging the volume and cost of technical debt unless organizations implement guardrails that balance innovation speed with technical validation.

The Illusion of Velocity: When AI Outpaces Governance

AI-generated code isn’t inherently flawed—it’s just unvalidated at sufficient speed and scale.

Consider the data: all LLMs exhibit model loss (hallucination). A recent research paper    assessing the quality of code generation of GitHub Copilot found an error rate of 20%. Compounding the issue is the sheer volume of AI output. A single developer can use a LLM to generate 10,000 lines of code in minutes, outpacing the ability of human developers to optimize and validate it. Legacy static analyzers, designed for human-written logic, struggle with the probabilistic patterns of AI outputs. The result? Bloated cloud bills from inefficient algorithms, compliance risks from unvetted dependencies, and critical failures lurking in production environments.

Our communities, companies and critical infrastructure all depend on scalable, sustainable and secure software. AI-driven technical debt seeping into the enterprise could mean business critical risk… or worse.

Reclaiming Control Without Killing the Vibe

The solution is not to abandon Generative AI for coding—it’s for developers to also deploy agentic AI systems as massively-scalable code optimizers and validators. An agentic model can use techniques like evolutionary algorithms to iteratively refine code across multiple LLMs to optimize it for key performance metrics — such as efficiency, runtime speed, memory usage –   and validate its performance and reliability under different conditions.

Three principles will separate enterprises who thrive with AI from those who will drown in AI-driven technical debt:

Scalable Validation is Non-Negotiable: Enterprises must adopt agentic AI systems capable of validating and optimizing AI-generated code at scale. Traditional manual reviews and legacy tools are insufficient to handle the volume and complexity of code produced by LLMs. Without scalable validation, inefficiencies, security vulnerabilities, and compliance risks will proliferate, eroding business value.

Balance Speed with Governance: While AI accelerates code production, governance frameworks must evolve to keep pace. Organizations need to implement guardrails that ensure AI-generated code meets quality, security, and performance standards without stifling innovation. This balance is critical to prevent the illusion of velocity from turning into a costly reality of technical debt.

Only AI Can Keep Up with AI: The sheer volume and complexity of AI-generated code demand equally advanced solutions. Enterprises must adopt AI-driven systems that can continuously analyze,optimize, and validate code at scale. These systems ensure that the speed of AI-powered development doesn’t compromise quality, security, or performance, enabling sustainable innovation without accruing crippling technical debt.

Vibe Coding: Let’s Not Get Carried Away

Enterprises that defer action on “vibe coding” will at some point have to face the music: margin erosion from runaway cloud costs, innovation paralysis as teams struggle to debug brittle code, mounting technical debt, and hidden risks of AI-introduced security flaws.

The path forward for developers and enterprise alike requires acknowledging that only AI can optimize and validate AI at scale. By giving developers access to agentic validation tools, they are free to embrace “vibe coding” without surrendering the enterprise to mounting AI-generated technical debt. As Karpathy notes, the potential of AI-generated code is exciting – even intoxicating. But in enterprise development, there must first be a vibe check by a new evolutionary breed of agentic AI.

[bunchacrunchcake]

YES YES YES

MAKE CLAPTRAP THE NEW AI MASCOT

[starryarchitect]

my hatred of ai born from being in creator communities and being a try-hard student who cares about learning more than grades and hates cheating of all kinds vs. my fear of not keeping up with a changing world and not adapting to new technologies and becoming the modern equivalent of my grandma who can't send an email FIGHT

[mytortoisegotstuckinatupperware]

*through gritted teeth* if Robert Langdon can grade student essays before going off to be the protagonist of angels and demons then so can I

[wahoo-stomp]

I spent five years coming up with unique ways to photograph the same group of plushies to help tell a story.

You don't need AI to help you be creative, you're just being lazy and want brain chemicals without doing any of the work or respecting the people who put time and effort into it.