106,838 views
Ralph Caspers explains what artificial intelligence can really do. Artificial intelligence draws amazingly good pictures based on text instructions and writes texts that are almost indistinguishable from "real" ones, i.e. those written by people. New, impressive programs are constantly being released. On the other hand, artificial intelligence sometimes makes mistakes that are so unbelievably wrong that they would never happen to a human. Where does that come from? The principle of today's artificial intelligence is very, very roughly copied from our nerve cells. So if a computer system is based on the network of human nerve cells and can actually learn, then you have to ask yourself: Why does it make such serious mistakes? And the other way around: Can artificial intelligence be more intelligent than humans? Are we threatened by a machine apocalypse like in "Terminator" or "The Matrix", in which machines take over the world? Ralph shows you the strengths and weaknesses of artificial intelligence in this video! A few AI programs to try out for yourself: https://www.craiyon.com/ https://www.wombo.art/create https://openai.com/dall-e-2/ https://chat.openai.com/chat Chapters: 00:00 This is what AI can do! 02:19 How does AI work? 06:48 Is AI really intelligent? 08:01 AI can be so stupid 10:12 Only humans can do that Our sources: Brief history of neural networks (English) / brief-history-of-neural-networks Why learning AIs are so easy to deceive: Worth reading article from the journal Nature on current AI: https://doi.org/10.1038/d41586-019-03... "Artificial general intelligence" will not come as soon as you might think: https://www.scientificamerican.com/ar... Dear Elon Musk, here are five things you might want to consider about "artificial general intelligence": https://garymarcus.substack.com/p/dea... Horse rides astronaut - What almost everyone misunderstood about DALL-E and Google's image and why you still shouldn't believe the hype about AI: https://garymarcus.substack.com/p/hor... AI madness: The strange side of machine learning: https://www.aiweirdness.com/ The Google developer who believes the company's AI has come to life AI ethicists have warned Google against imitating humans. Now one of Google's own employees believes there's a ghost inside the machine: https://www.washingtonpost.com/techno... Authors: Ralph Caspers, Ingo Knopf Editing / Graphics: Klaus Wache Sound design: Klaus Wache Did you like the video? Then subscribe now to the official Quarks channel with Ralph Caspers @DimensionRalph on YouTube! Also visit the official Quarks channel on YouTube: @Quarks #Quarks on Instagram: / quarks.de / beautyquarks Quarks on Twitter: / quarkswdr Quarks on Facebook: / quarks.de #ArtificialIntelligence #dimensionralph #ralphcaspers #AI