Recent years have seen a lot of speculation about both the potential and limitations of art caused by artificial intelligence. Technicians and novices alike look to bring out tools like deep dream And the DALL-E, discuss their aesthetic advantages, and ask what programs based on similar algorithms might achieve in the future. The new documentary computer accent It repeats many of the talking points in these discussions without adding much to it, but makes one useful contribution to the topic: an extended, in-depth look at the actual process of creating art using artificial intelligence.
In 2016, the pop dance group YACHT – musicians Claire L. Evans and Jonah Bechtolt – decided to use artificial intelligence to help compose their next album. Directors Reel Roche-Dicter and their crew have followed the band for several years as they pursue this vision, learning about neural networks and devising strategies for how to devise a way to extract usable music from an instrument. Eventually, Evans, Bechtolt and their technology and music collaborators transcribe the entire backend catalog of YACHT into MIDI, which is then fed into MusicVAE, a tool from Magenta (Google’s open source project for machine learning research). With the model “trained” on this information, it can then create its own MIDI snippets based on pre-existing YACHT music. The band takes what they can from that output — the tracks that interest them, the clips that sound good — and then perfects and combines the material into different songs. The result is their 2019 album Series kick off.
Frustrating , computer accent, much like the discourse on artificial intelligence and art, operates from a misunderstanding of its subject. Among these misunderstandings is the idea that any of these tools really constitute “artificial intelligence”. Programs based on algorithms are nothing more than highly complex complexes. While the art they can produce increasingly looks like human-made art as technology has been improved, this is not intelligence getting “smarter” but a more complex form of imitation.
The general assumption is that feeding more information into these programs will help them better approximate the human mind, as if cognition could be quantified. But pattern recognition and association are components of sensation, not its foundation. These traits by themselves do not build on complex thinking; They form prejudices and stereotypes (as evidenced by AI’s persistent problem with perpetuating racism). early in the computer accent, one expert warns against embodying these tools. However, even if Evans and Bechtolt only refer to MusicVAE as a “collaborator” in the humor, the film still assumes man and machine in those terms.
That’s a shame because it ignores what’s really interesting about what YACHT does chain kicks offIt is a new songwriting technique. The idea of an “AI-compiled album” may inspire pictures of a computer spitting out music, but the process requires a tremendous amount of human organization and redistribution. However, the results mostly look like YACHT’s rejuvenation of familiar territory…because, of course, all an AI can currently do is remix what’s put into it. However, the YACHT trial was well worth a try, even if its results were mixed. As a document of that experience, computer accent Helps clarify potential uses for AI, as well as how much Much Well you should go.
computer accent Currently in theaters.