AI and Hollywood
AI infringement upon performers’ rights is an interesting topic, but I believe that most of the current debate will soon become moot.
When CGI can create photorealistic scenes, and AI can string these scenes together into video that is indistinguishable from professional film-making (at least, to the layman), then the limiting factor becomes the capacity of AIs to turn well-written scripts into well-performed movies and television.
As with other cases of technological unemployment, this will create even more opportunities for superstars while crowding out other performers.
Much like with writing, video “stars” are recognized as such by virtue of having established strong platforms, reputations or followings.
Once the copyright issues are sorted out, which I am confident will redound to the benefit of the performers, superstars will selectively license use of their images into AI-generated video productions and enable their estates to do so, when they are deceased.
This will make it far more challenging for new artists–in all creative media, not just video–to break into what are already extremely rarified professions.
By the 2030s, I would not be surprised to see producers offering viewers their choice of actors for a given role, from a set of licensed image and voice attributes that define each “actor” for copyright purposes. Each such “actor” will then receive a “per-performance” fee, and probably also a one-time “right to use” fee. (Would you like for the lead role in the mystery film you are about to watch be played by Humphrey Bogart, Marlon Brando, Leonardo Dicaprio, or Robert DeNiro? Click on the name to see a trailer featuring that actor.)
The huge numbers of people behind the scenes who participate in the making of a studio film or television episode (just watch all of the credits to appreciate how many are involved), will also lose most of their employment opportunities. However, in these cases, when the human expertise of superstars is used to add an extra “wow” factor to the finished product, the public will likely never know the names of the professionals who are involved.
One of our AI expert allies advises that, eventually, we can anticipate studios creating entirely synthetic superstars. How? By creating “avatars”; each with a distinctive face, mannerisms, and voice, and then using such avatars for roles in movies and television. Instead of having fifty takes of a scene with human actors, the director can have tech people tweak the delivery and especially the emotional reaction in the synthetic performance. So, eventually even human superstars may not be marketable–though I expect that those of us who have favorite actors will continue to want to watch them, so this development should be decades away.
All of this will enhance consumer options while reducing prices thanks to the slashing of production costs. It will be yet another case of technological unemployment, among a rapidly growing set of such cases–all of them increasing superstars’ incomes and corporate profits, while reducing consumer prices and, sadly, leaving non-superstars without an opportunity to earn incomes in their preferred professions.
Ultimately, this will prove to be further evidence that societies need a universal basic income that is both viable and sustainable. The only such proposal of which I am aware is the MOUBI from Basic Income Australia, conceived independently by both our ally Michael Haines and Robert Heinlein (!)
Some, who are outraged by all of this, will demand that studios never use AIs to produce films and television. Should such indignation ever gain real traction in the marketplace–and we must never underestimate our human capacity for shortsightedness and foolishness–the main effect will be to shift video production from countries that restrict it to countries that don’t. (Consider the US’ shortsighted attempts to ban stem cell research early in this century. The main effect was an exodus of talent to more receptive nations.)
Any attempt at such regulation or law must contend with many thorny issues. Among them:
- What differentiates AI from other software? (There is zero chance of banning software use in general, and our AI expert allies advise that it is almost impossible to draw what attorneys call a bright line between AI and other software. Trying to rule that specific uses of AI are valid will be like wrestling with the proverbial genie, who provides what was requested but rarely what was wanted.)
- Should use of AI be banned in the making of commercial film and television, what will prevent the use of same by “amateurs” on Youtube and elsewhere, whose production quality will rapidly approach that of professionals, thanks to AI?
- Given the complete long-term failure of previous historical efforts by Luddites to block technological improvement, how can laws and/or regulations work this time? (Tip: they can’t.)
Any attempts to legislate such shortsightedness will create many more jobs–for lawyers; and later for legal AIs.