glorified autocomplete. profit-seeking trojan horse for propaganda purposes. self-inflicted workforce deskilling. unsustainability generator.

ai is shit

Thank you for coming to my ted talk.

It's upsetting because linguistically I generally find anthropomorphization a lot of fun. Tree struck by lightning? Poor guy hasn't been having a good time. Microwave sometimes unresponsive when you press the buttons? He's been a little disagreeable lately.

I don't want to give the AI dignity, but allowing it to change my speech patterns also seems like dignifying it.

I liked some proposals. I will certainly use LLM far more, and i/o too. But that's me, I have a preference toward technical terms that are already established. Finding new ones is a challenge. Specialy since they get permeated by our own biases. Or at least that's what I felt when reading their proposal to replace ai agent. I believe the agency of that system is there, so why not just llm agent? :/

Examples from the article:

• artificial intelligence → probabilistic automation
• hybrid intelligence → augmented human intelligence
• image recognition → image labeling
• speech recognition → automatic transcription
• the model shows bias → the model reflects bias
• model mistakes → model errors
• chatbots are good at … → chatbots are good for …
• hallucination → undesirable output
• goal → success condition