> to make up answers [...] to make up APIs and lower my code quality [...] to answer customer questions and make up company policies [...] to miscategorize imag

> to make up answers [...] to make up APIs and lower my code quality [...] to answer customer questions and make up company policies [...] to miscategorize images These are valid weaknesses of current tech, but most likely to just improve with time. Networks are already above human-level on many image categorization tasks. > to produce generic art for unimaginative creators Personally I think novelty in art thrives when the entry barrier is low and experimentation is possible without a huge investment, which AI can help achieve. A lot of works/trends I've found interesting recently have involved some creative usage of AI (e.g: https://i.imgur.com/Jybvj0r.png https://i.imgur.com/Jybvj0r.png - zoom out). Granted, Sturgeon's law is still in full effect. > to pilot autonomous vehicles that injure pedestrians and take jobs from taxi drivers and truckers. When we get to the point where they injure pedestrians at a lower rate than human drivers, then yes. We should be fighting for UBI or broader economic changes to ensure everyone (not just rightsholders like Getty/Universal) can benefit, not trying to ensure that human labor will always be required to drive trucks across the country. > AI that can't taste to create recipes [...] that's never traveled to plan my vacation [...] that can't appreciate music to recommend music for me If I'm actively seeking a recommendation, I don't see an inherent issue with that being done through statistical methods that take into account the preferences of many. > I need AI to replace web search [...] Web search has been AI for either years (using machine learning with transformers/language model embeddings) or since it's inception, depending on your definition of AI. Same for malware scanning, spam filtering, DDoS prevention, reverse image search, language translation, fraud prevention, material/product defect detection, voice dictation, drug discovery and interaction prediction, logistics optimization, infectious disease modelling, protein folding, weather forecasting/early-warning systems, tumour segmentation, video transcription, and so on.