Gregory's Server@iolo Until we understand what AI is doing and how it works, so we understand the limitations, we should be careful about how we use it.
If you've got a vague understanding about how the generative AIs work, you'll know why it produced this image.
|
6 comments
@zdl @iolo I completely agree. And what's worse is investors pumping vast amounts of money into generative AI without understanding the basics of how it does what it does. Or they do understand, but they're just riding the wave out of greed. At least the NFT bubble only lost money. This will cost jobs, devalue all information and yes, potentially cost lives.  @guigsy @zdl @iolo whats far more worrying is there seem to be many business owners/executives who are at least smart enough to remove the obvious crap/hallucinations, but are already using AI art to eliminate jobs for entry level graphic designers, content writers and even human models (saw an advert for mobile phone / fitness tracker kit and the unusually perfect looking "mixed race" lass modelling them looks suspiciously like she is AI generated)  @vfrmedia @zdl @iolo replacing artists, actors and models with AI is a moral choice that's influenced by cost. It's a different issue from presenting information that's factually incorrect. And that there should be AI tools to generate information that it's likely to be factually flawed because we know that it will be published without being validated. the risk of misidentifying poison mushrooms due to AI is a real danger. Miswired electric wiring schemes less so, as there is much more reliable information on these available by conventional means (in UK we are often taught the correct wiring colours in high school). Its going to be a further dilemma for Communications Ministries worldwide to balance allowing free expression with clamping down on dangerous misinformation (from AI or any other source) |
@guigsy @iolo There are about 8 billion people on the planet, give or take. What proportion of them are in a position to understand AI to the point that they even know there *are* limitations, especially given the intensely dishonest marketing its proponents push it with?
People are going to die over this foul technology, whether from eating toxic mushrooms because a book someone "wrote" with degenerative AI hallucinated toxic mushrooms into edible ones, or because people wire up plugs wrong.