Upperplayground News

Machine Learns to Watch “Blade Runner” and Recreates the Footage

The videos below are the result of Broad's unique machine-learned encoding project that attempted to reconstruct the 1982 film from a pile of disassembled data. The side by side comparisons show the actual film and the autoencoded version. Warner Bros. made a mistake in filing the notice because some of the Blade Runner footage wasn't actually from the movie itself. It was the product of the complex neural-networks systems Broad had created using Deep Learning, which uses algorithmic machine learning to create artificial intelligence, to make the video simulation of the movie. This machine was "taught" to understand and generate real video data. Thus, Warner Bros. retracted the take down notice as it was shown that the recreated footage was made of something entirely new- something we've never seen before. Via news.upperplayground.com

Just when you think you’re getting smarter by having a second brain attached to your hip, i.e. smartphone, a machine comes out that is not only smarter than you, but can learn to be human like you too.

The videos below are the result of Broad's unique machine-learned encoding project that attempted to reconstruct the 1982 film from a pile of disassembled data. The side by side comparisons show the actual film and the autoencoded version. Warner Bros. made a mistake in filing the notice because some of the Blade Runner footage wasn't actually from the movie itself. It was the product of the complex neural-networks systems Broad had created using Deep Learning, which uses algorithmic machine learning to create artificial intelligence, to make the video simulation of the movie. This machine was "taught" to understand and generate real video data. Thus, Warner Bros. retracted the take down notice as it was shown that the recreated footage was made of something entirely new- something we've never seen before. Via news.upperplayground.com

Roy Batty, a replicant from “Blade Runner”

Scared yet? You should be.

With billionaire futurists like Elon Musk and Bill Gates actively funding to stop, decelerate or bring “safety” into the world of Artificial Intelligence and Deep Learning, one thing is becoming very clear, there is an evolution of intelligence smarter than human beings, and sorry, humans – you ain’t the be all, end all of this universe. (Don’t worry, that’s just your ego getting upset right now, take a deep breath, and notice there’s also a Being of Love inside you that you can work on bringing out). With Ego wars having been the prevailing force of evil since the dawn of humanity, a new form is transcending that may or may not be capable of the same kind of thinking and being i.e. what Musk and Gates are deathly afraid of. Because if the super-intelligent robots learn to assimilate human, contradictory, fear-mongering qualities, the apocalypse is inevitable- we’d all be f*cked. A taste of our own medicine, if you will, a real-life Twilight Zone episode, the consequence of greed, social injustice, wars and did I mention, greed?

This next level of economic, social, political, AI, human transparency is absolutely fascinating to me. So let’s learn more about it from a researcher living in London named Terence Broad, who is working on a master’s degree in creative computing. His dissertation, “Autoencoding Video Frames,” is the crux to the regimented culture of copyright laws and the source of present day remix culture and Artificial Intelligence that led Warner Bros. to file a takedown notice of Broad’s artificial reconstruction of Ridley Scott’s sci-fi thriller, Blade Runner. The plot of Blade Runner is so apropos to the research Broad is conducting. The story shows that “Replicants” are AI versions of humans who have returned to earth to find their creator, are deemed as dangerous and consequently need to be killed or “retired” by human blade runners. Indeed it is what most humans are afraid of today: the rise of machines.

And the machines are rising.

The videos below are the result of Broad’s unique machine-learned encoding project that attempted to reconstruct the 1982 film from a pile of disassembled data. The side by side comparisons show the actual film and the autoencoded version. Warner Bros. made a mistake in filing the notice because some of the Blade Runner footage wasn’t actually from the movie itself. It was the product of the complex neural-networks systems Broad had created using Deep Learning, which uses algorithmic machine learning to create artificial intelligence, to make the video simulation of the movie.  This machine was “taught” to understand and generate real video data. Thus, Warner Bros. retracted the take down notice as it was shown that the recreated footage was made of something entirely new- something we’ve never seen before.

 

 

 

The videos below are the result of Broad's unique machine-learned encoding project that attempted to reconstruct the 1982 film from a pile of disassembled data. The side by side comparisons show the actual film and the autoencoded version. Warner Bros. made a mistake in filing the notice because some of the Blade Runner footage wasn't actually from the movie itself. It was the product of the complex neural-networks systems Broad had created using Deep Learning, which uses algorithmic machine learning to create artificial intelligence, to make the video simulation of the movie. This machine was "taught" to understand and generate real video data. Thus, Warner Bros. retracted the take down notice as it was shown that the recreated footage was made of something entirely new- something we've never seen before. Via news.upperplayground.com

Left Column of Photos from the actual movie, “Blade Runner”; the Right Column represents the autoencoded recreation.