Aardvark Daily

Please visit the sponsor!

Sora changes everything, including copyright

"I want to see video or it never happened" is the catchphrase often heard these days.

In an era where everyone walks around with a 4K camera in their pocket, thanks to modern smartphones, we get to see videos of almost everything that happens in the world. Even the most unlikely events are now frequently caught and immortalised in video then distributed around the world by the internet.

Other people's lenses and cameras help shape our view of the world and expose us to things we could only otherwise dream of seeing.

Until now.

Artificial intelligence has taken another leap forward with the launch of Sora by OpenAI.

Now we can no longer trust our eyes when we watch video of some earth-shattering event or unlikely situation -- because it may not be real.

Skilled VFX artists have, for decades, been creating illusions in video and this has become a part of almost every blockbuster movie today. However, creating these sophisticated visual sequences is a slow and expensive process that requires highly skilled artists and expensive computer systems. Well that was the case until last week.

OpenAI's Sora is a generative AI system that, despite being an early release, already creates stunningly realistic video sequences based on text prompts.

This Arstechnica story gives some pretty impressive sample footage.

Now you can create video of *anything* your imagination can think of.

That is both impressive and scary, at the same time.

I shudder to think how many "creatives" will lose their jobs once this becomes more mainstream.

Why hire a video crew, VFX artists, compositors, drone pilots and the vast array of other professionals that are now needed on even the smallest production, such as a TV ad, when some carefuly crafted text commands can produce possibly superior results in a few minutes and for what will likely be a few tens of dollars?

Perhaps the only hurdle standing in the way of the widespread commercial acceptance of this tech is the thorny issue of copyright.

As things presently stand, a work created by AI can not be copyrighted (if I understand things correctly). Due to this, anyone who creates content using AI will have a great deal of difficulty protecting that content from re-use by others. If you can't protect your content from re-use, it becomes far more difficult to make a fortune from it.

Perhaps this is why we've yet to see a top-10 tune created by AI -- there'd just be no way for music studios to squeeze blood out of that AI stone because anyone could simply and freely copy and redistribute that work without fear of infringing copyright.

Now that we've reached a point where generative AI truly can create works with commercial value I would expect to see the recording companies, movie studios and other "influential" parties successfully lobbying (ie: buying) governments to change copyright laws so as to enable profits to be made.

Bookmark this column and come back in a year or so -- by which time those lobbying efforts will already be well under way and possibly even have succeeded. Governments never move quite as quickly as when funded by huge corporations.

In the meantime, don't believe everything you see on the internet!

Carpe Diem folks!

Please visit the sponsor!

PERMALINK to this column

Rank This Aardvark Page

Change Font