For a long time now, cats have known what a computer is. Why else would they prowl all over our keyboards? Computers on the other hand, only learnt ago what a cat is a few days! Google’s latest trick is to automatically recognise not only human faces but also cats in images on the internet.
The revolutionary aspect is not that Facebook and other social media platforms are going to allow users to tag cats online, but that computers taught themselves to recognise cats! That’s the hot topic! Whilst it’s still being argued what A.I. (artificial intelligence) can and may do for people, there is almost unanimous agreement that self-learning software will provide innovations on an unimaginable scale.
What applies to photos also applies to moving pictures
With sufficient capacity, that which works for a single image will also work for 30 images (frames) per second. It’s simply a question of time until computers will be able to automatically analyse and catalogue films and videos. Monitoring software for security purposes can do this already for faces, and in real time too. It’s enough to simply count down from five and turn around for our video editing software will tell us what we’re seeing! This is useful for archiving purposes and practical for film databases, but that’s not all it’s good for…
1. Image recognition becomes automated
With the exception of experimental works and music videos, video editing follows a set of rules rather than the story that’s being told. You don’t need to be a quantum computer to think that, from a mathematical point of view, the number of conventions that have to be observed (or deliberately ignored) when editing video is ridiculously small. I reckon that any internet search algorithm is more complex and requires more brainpower and storage space then the rules and conventions of video editing software!
It’s simply a question of time until computers will be able to automatically produce films and videos.videothink
Considering the fact that software will be able to learn by itself in the future and thus develop itself independently, things will get really exciting in the postproduction stage. An enormous amount of cinematic work of all genres has already been digitalised and is available online in varying qualities – from cat videos on YouTube to Hollywood blockbusters starring Tom Cruise. Adobe’s Premiere initially asked for genre and length. A few versions later and the software began making footage suggestions immediately after reading the filmed material. But computers don’t just stop learning.
A little later the editing program also started taking into account things it had learnt from digital distribution. Finally, the computer knew who clicked away when during the film. For a small fee, users can receive a film cut specifically to their personal tastes, which corresponds to their online digital footprint, and takes their personal, private preferences and dislikes into account. Paradoxically, every blockbuster would then be technically different specifically in order to have the same psychological effect for each consumer.
2. Acquiring content will happen virtually
Both Buster Keaton and Charles Spencer Chaplin knew that anything that wasn’t convincing in front of the camera would not be improved by being recorded onto celluloid. In the future, this rule will be thrown out the window. Today, feature films are already being animated on computers without a one single, genuine day’s shooting. Video editing computers will be able to add one and one and generate the perfect material for the perfect film. Disney says that in 5 years, digitally animated images of people will look no different from the real thing. I assume that representatives of a US company listed on the stock exchange are cautious with such statements meaning the reality is probably closer to three years. This will revolutionise films and videos.
Conclusion and final question
Computers have been able to create texts since they were invented. Therefore, in this brave new world, the screenplay will be loaded into the computer in place of raw material. The software will then supply a finished film in the desired resolution, tailored to the user via the internet. However, will people still write the screenplays and will they even watch the films? By then cats won’t be stalking over our keyboards, instead they’ll be connected to the internet themselves via wifi. They’ll look like cats, behave like cats and will be as much like cats as the seals that nowadays entertain the residents of Japanese retirement homes.
Read more about:
Superintelligence: Paths, Dangers, Strategies, written by Nick Bostrom,
published by Oxford University Press