The phrase “AI” has been poisoned so hard that nobody building serious products should use it. I would love to see more free software projects leverage computer vision or natural language processing for example, but I would never describe those as “AI” because it’s a brand destroying phrase at this point
AI is the failed pin that can’t actually do anything. AI is lying chat bots. AI is crappier search. AI is stealing other people’s art. AI is enshittification. It’s a bad look and we should steer clear of that phrase. /hottake
@danirabbit
AI is investment hype, it's the latest in a series of "gold rushes". I made a big long post about that no one has read, but that's the gist of it.
@felichsdakatze hey, linky please! I'd like to read it. I've been through this cycle more than a few times and each time it makes me want to shove a barbecued shoe down my throat.
The worst is when you're actually working on something legitimate that did fit the hot word of the day and now its identity has been completely redefined out from under it.
If it feels familiar, that's because it is. AI isn't about creating a technical solution or product. It's about something that looks like it could be either of those things to hype up the market. Same as wework, same as self driving cars. If they've lightened your pockets along the way, they've achieved their goals.
Tech bewilders those with spare change to invest. And those people mostly have a poor understanding of tech.
@danirabbit No joke. At this point, most #marketing copy using the terms "AI" or "artificial intelligence" really just means "uses computer."
For real, my security camera is not using #AI to find humans in the frame. That's basic #ImageProcessing that's been around for years. My #DSLR could accurately fimd and focus on faces a decade ago. That certainly was not #ArtificialIntelligence!
@danirabbit I for one think many associate artificial intelligence with general artificial intelligence, which nowadays technology is nowhere near. Therefore, using fitting descriptions of subfields such as machine- or deep learning instead might be less ambiguously and sensationalistic.
Thus, an open source application employing a voice recogniser is not "an AI assistant" but simply an application deploying deep learning techniques.
Advanced explanations might also be appropriate.
Add comment