@mikarv the illusion of productivity is one hell of a drug

@MissConstrue this is horrific... I'm so sorry for that city's loss

@stevenaleach yeah this is why I can't understand the moral panic against deploying text or image classifier neural networks to detect synthetic content. it may feel like an unwelcome capitulation to "AI" for those who are strongly opposed, but the alternatives are far, far worse

@xgranade @violetmadder @kahomono Probably the case that anti-AI absolutists might be sympathetic with is the software engineer who is being literally forced by their management to use AI. For them, the least harmful pattern is: self-host the StarCoder 3B and Apertus-8B LLMs with Ollama on a local machine, and configure the Continue extension for VSCodium to use StarCoder for autocomplete and Apertus for interactive chat. Also, use solar power.

@xgranade @violetmadder @kahomono There are less harmful tools that fall into the category of things commonly labeled as AI, and less harmful ways to use them, but those patterns tend to be more limited in scope, not all-in-one, one stop shop apps or websites.

@xgranade @violetmadder @kahomono I am going to regret sticking my neck out on this thread. but, I have been wrestling with your question for a long time, and I think that, along with self-hosting (while sourcing your own power more sustainably) and striving to use models that have been trained with attention to human dignity, a key thing to push back on is the cult of AGI, an AI that does everything.

If there's interest, I'll share my little ChatGPT/Claude/Gemini/DeepSeek clone that wasn't trained on stolen data and doesn't require massive energy usage. Honestly, it was largely vibe-coded anyway (yes I used Claude to build its own replacement). But if nobody cares, I'll just keep using it for the things I'm using it for, and skip the emotional labor of cleaning it up and putting it out there to be attacked from all sides. At least I won't feel quite so guilty about using AI at all. (end 🧵)