Re:Ask HN: Is the AI Apocalypse Imminent?
by 082349872349872 on Friday December 09, 2022 @03:09AM
(#33918878)
Compared to the early days of the net (4 decades ago), there's already an amazing amount of low-quality generated content that is being uploaded just for the numbers. Will in-silico generation make that worse than wetware generation? At the moment I'm somewhat optimistic that as the machine generated content now has a better grasp of spelling and grammar, it may eventually prove more coherent than any of the least-coherent wetware sources.
In the best case, people with good ideas but poor articulation could use AI (as a far far cheaper substitute for wetware legal counsel) to put their arguments in succinct and lucid forms.
Example: AI can now generate great single concept arts, but in my opinion it will still take some time until it can do it coherently for a full project where everything needs to fit together. In the same manner the developer needs to write code that fits into existing systems. Both can of course profit already from AI today, but they are not to be replaced as easily.
The way bigger threat lies in all the social aspects of the internet. It's hard already to weed out all the crap when I want to find something specific e.g. on Youtube. I imagine it will be even harder when I need to filter through the low-quality generated content that will be uploaded just for the numbers. Also I see non-curated online discussion platforms and comment sections dying: How am I supposed to properly discuss when everytime I take a stance there will be instantly 10 bots screaming back at me?