Scientists warn of AI collapse

in the past year or so we've all become used to AI generated text and images and audio and increasingly also videos there's been a lot of talk about how terrible this is for writers and artists and so on but some computer scientists are warning that this AI creativity May soon collapse let's have a look the problem is fairly easy to understand but difficult to quantify the AIS that we currently use are deep neural networks that are fed huge amounts of data and basically learn to recognize and reproduce patterns large language models recognize grammatical rules and words that belong to each other image creation software recognizes shapes and shadows and gradients video software recognizes moving shapes and their context and so on but where does that data come from that they need to learn well that was created by the original neural networks humans the issue is now that the more people use AIS to create new content the higher the risk that future AIS will be fed data that they have produced themselves and what will this do it's not a priority all that obvious you might think that with AI having a random element and sometimes being prone to generate nonsense the result might be that it just produces increasingly weird stuff but actually the opposite seems to be the case both for language and images the more AI eats its own output the less variety the output has for example in a paper from November a group of scientists from France tested this for a large language model they used an open- Source model called opt from meta and developed several measures for the diversity of language then they tested what happens for the diversity of language for tasks requiring different levels of creativity for example summarizing a news article requires low creativity writing a story from a prompt requires High creativity in this table they summarize the language diversity scores for the levels of training iteration as you can see they pretty much all drop the language diversity drop is especially rapid for storytelling a similar finding was made earlier by a group from Japan for AI generated images based on stable diffusion the AIS decrease the diversity of the image set and if you train them on their own output diversity continues to decrease you can see this rather clearly in the image sets that they use as examples these are some examples of real elephant images from the original data set that they use and these are some examples of the images that the AI generated after training as you can see they have some of the familiar problems some legs too many or too few or to heads and some conflation of body parts but the most striking thing is if you look at a collage on the left is the sample of the original images on the right the AI generated ones you see immediately that the AI generated ones are much more alike I think that many of us have by now noticed that if you've been using mid Journey for some while you have learned to recognize mid journey is images even leaving aside the obvious problems that these images continue to have they tend to Output similar looking images for example unless otherwise instructed people tend to be white young and good-looking these are four images that mid Journey created when prompted with human face photo realistic without further instructions as you can see they all look more or less the same what are the consequences well no one really knows the issue is that our entire environment is basically being contaminated by AI generated content and since there's no way to identify its origin it'll inevitably leak into training data it's like plastic pollution it won't be long until we all eat and breathe the stuff there are two ways things can go from here one is that it turns out that this is a general problem which can't can't be overcome with these types of models in which case well good news for humans our creativity will still be needed it also seems likely to me that AI generated content will have to be marked as such I suspect that this is where laws will take us the other way it could go is that the next generation of AI will remedy this problem by deliberately enforcing variety for example by making more use of Randomness and that we'll simply give up trying to distinguish AI generated content from Human generated content what do you think let me know in the comments if you want to learn more about how neural networks work I recommend you check out the neural network course on brilliant.org who've been sponsoring this video the neuron Network course will give you a deeper understanding of how intelligent artificial intelligence really is with some handson examples and Brilliant has caused us on many other topics in Science and Mathematics too whether whether you're interested in neuronet or Quantum Computing or linear algebra they have you covered I even have my own course there that's an introduction to Quantum Mechanics it'll bring you up to speed on all the basics interference superpositions entanglement and up to the uncertainty principle and bells theorem brilliant is really the best place to build up your background knowledge on all those science videos which you've been watching you can try it out for free for 30 days but if you go there use our link brilliant.org Saina because the first 200 to use our link will get 20% off the annual premium subscription so go and give it a try brilliant is time well spent thanks for watching see you tomorrow

Channel: Sabine Hossenfelder

Views: 787,249

Keywords: science without the gobbledygoook, hossenfelder, science news sabine, science humor, science news, tech, tech news, technology, technology news, ai, ai news, ai creativity, ai creativity vs human creativity, future of ai, ai taking over the world, chatgpt, will ai replace artists, ai vs human, large language models, AI collapse, ai model collapse

Id: NcH7fHtqGYM

Length: 5min 49sec (349 seconds)

Published: Mon Mar 04 2024