What happens when programmers get to vote on a complete ban on slopcoded software on their code platform? Members of Codeberg were asked to vote on a proposal to completely ban slopcoded projects from Codeberg, and in what should not be a surprising outcome to anyone not overcome with “AI” hysteria, the vast majority voted in favour of the complete ban: over 70% of Codeberg members voted to ban slopcoded projects entirely (358 in favour, 144 against, 14 abstentions).
Of course, this is not a surprising outcome. Stripped down, programming is a form of artistic expression, and programming is no different than writing, painting, composing, or any other artistic endeavour. I think it’s safe to say writers, painters, composers, and similar creatives are against “AI”, so it’s only natural programmers feel the same; poll after poll shows the overwhelming majority of respondents – usually well over 70-80% – are against “AI”.
The pro-“AI” accounts on OSNews often try to paint my anti-“AI” position as extremist, but in reality, I’m just voicing how 70-80% of people clearly state they feel. I have zero skin in this game, zero outside pressure to please any bosses to get that promotion, zero pressure to conform to avoid getting laid off, zero pressure to not contradict upper-management. I can speak freely, openly, and without fear of retaliation. And as Nikhil Suresh explains in his harrowing from-the-trenches article AI Mania Is Eviscerating Global Decision-Making, that’s a massive asset.
The vast majority of people – including your friends, family, and co-workers – really hate “AI”. You can either accept this, or be left behind.
Reading the comments, the obvious reason for it was provided, AI produced code is heavily affected by copyright infringement and hence it represents a liability to host it, use it, redistribute it, sharing it … Even most open source licensees at minimum demand author attribution. And we all know, as it was publicly admitted, everything in existence and regardless of the license was thrown into it and now AI is sharing it with all original conditions stripped away.
Thom Holwerda,
This is clearly about my recent post. I don’t think you realize just how popular AI assistants are becoming with ordinary users who don’t have a horse in the race.
The 70% vote in the article was for the following amendment, which doesn’t actually ban AI coding assistants. It only bans projects consisting of mostly generated code. It’s a significant difference that AI helpers are still allowed.
I’m not really bothered by the policy. Ban projects generated without much thought and without an author who understands the code…I’m fine with that. But this doesn’t necessarily mean those 70% are against the use AI assistants altogether. If a developer works with AI to solve a problem, then it’s allowed by the policy and I don’t have a problem with this either.
In reality things are much more nuanced that you are admitting though. Most people are FOR AI assistants that answer questions and help them do tasks, find bugs, but against AI competing for resources, don’t want data centers in their backyards, AI spam and slop videos, and so on. In short there are good and bad aspects of AI, a single value is just not going to fairly reflect this nuance. I for one hate some of these bad AI use cases more than I like good AI uses cases. So if someone asked me if AI is good or bad, well the question itself is inadequate: AI is actually both.
I know you’re unlikely to change your ways, but I do think there are problems with your coverage that should be addressed. Like putting quotes around “AI” and pretending that context doesn’t matter when you trash it. Now it may well be your opinion that all AI is bad (you’ve refused to clarify), but I do think that is solidly into extremist territory and is not representative of how most people feel about AI. AI has so many applications that ordinary people would approve of, like cancer screening, helping the blind, helping the deaf, planning crop rotations, etc.
Of course it’s possible your opinion is actually more nuanced and less extreme than your posts suggest. Maybe your actual criticism is for specific uses of AI and not everything under the sun. If so, that’s fine, I’m genuinely curious to hear more about that. However I’d put to you that if there is confusion over nuances of your opinion, your own words are to blame since you routinely summarily blame all of “AI” as a generic concept. If this is not what you mean to say, then why not actually say what you do mean? Use more specific words like LLMs when what you are criticizing is LLMs.
I’ve been reading this site for a very long time – I very rarely jump into these types of conversations, but I feel like I need to here – despite the fact that this has practically become a religious argument and I know it’s probably not going to change anyone’s mind.
I’m actually disappointed that you, and this website by extension, are so vehemently against what is a technological advancement – whether y0u personally like it or not. It smacks of being like the old mainframe grey-beard operator who was never moving to a client/server model, the Netware guy who was “moving to that Windows crap” over his dead body, the tech who wanted to run dedicated hardware 1Us and 2Us in a rack instead of moving to a virtualization platform, or the sysadmin who dismissed cloud services as “just someone else’s server” while desperately trying to maintain uptime on-premises. They all got left behind.
Is there a lot of hype in the “AI Space” right now? Yes. If I go to the grocery store, are my Lucky Charms and Cool Ranch Doritos now coming with AI? Probably. The same thing happens with every tech cycle that I’ve seen in my life. Then the hype goes away. Then the real work gets done. Then it just becomes what it was supposed to be all along – integrated into our computing workflows.
My opinion is that using these LLM’s/weak AIs is exactly where computing was supposed to go, as far as I’ve seen it throughout my life. It’s another step towards the ultimate “future of computing”, whatever that is. They’ve been telling me that this was 5-10 years away since I read about it in my “Weekly Reader” in first grade 50+ years ago – along with nuclear fusion and the cure for cancer. And it was *always* 5-10 years away, until ChatGPT landed and truly disrupted. Is this the final version of this stuff? Absolutely not. I’ve seen ideas resurface over and over again – the first iteration is too soon, the second iteration is usable but isn’t quite right, the third iteration is where it becomes truly viable and usable. Platforms came and went. “Cloud Storage” popped up twice before we hit OneDrive or iCloud. The tech industry is littered with examples of this across hardware, software, and the web/cloud/et cetera. We’re probably in the second iteration of this now, but it is evolving quickly.
I’m guessing that the majority of the readers on this site have seen an episode or two of Star Trek: The Next Generation. I don’t once recall watching Mr. Data throwing on a hoodie, jacking a laptop into the holodeck, and then hacking away on C++ code for hour upon hour to build a program. The computer was prompted. You are hung up on, from what I have read, that it’s trained on copyrighted materials – that it uses FOSS code without attribution – it’s using proprietary knowledge without permission – it’s not creating anything – it’s all slop and crap and garbage. And I agree that some of it is. But, some of it is not. Using LLMs/AI can certain automate and orchestrate routine tasks – it CAN be a creative endeavor if the person behind it understands exactly what he or she AND the LLM is actually doing.
I’ve said this over the years more times than I can count – about PCs in general, about the web, about the cloud – This is a bus – either you get on it or you get run over by it. You don’t have to like this transition, but if you work in technology or are a technology enthusiast, don’t you owe it to yourself and the people who will come after you to understand it well enough to help shape what comes next? The conversation that is worth having is not whether AI is pure, ethical, intelligent, creative, or worthy of the hype. The real conversation we need to be having is “How do we shape it into something useful while limiting the damage it can cause”? The stand we need to take isn’t calling it slop, garbage, or crap and wishing it would go away. It’s recognizing that this is a social, legal, economic, and technological transformation – and then insist that we all get a say into how it develops.
We need to demand that transparency, attribution, privacy, and accountability is built into this. We need to ensure that we implement protections against abuse. We need to determine where these tools work, where these tools actually belong, and establish the boundaries of their use. We need to expose them when they fail – and examine how and why they failed, and what we can do to improve them and keep them from failing again. Crying foul and not contributing constructively is just leaving the decisions to the people, companies, and governments that are willing to make them.
Seems like a reasonable stance by Codeberg to me. If you love AI you can use Github and everyone will be happy.