No it isn't. Just make a good movie. Forrest Gump didn't need any machine learning.
The only reason someone might use this is when your strategy is "spray and pray", create crappy movies that insult everyone's intelligence and you need to figure out is it worth launching at all.
No, it just needed a massive demographic the studios could exploit the Hell out of by taking a semi-popular novel and Spielberging all over it.
You have a good point, but you used an imperfect film to demonstrate it. Forrest Gump was just about as cynical a film as you could imagine, and succeeded precisely as predicted. That kind of cynical filmmaking, explicitly targeted at demographics with earnings projections in place before casting begins, is precisely what this AI-driven concept will end up with at its worst.
A better example would be the works of the Coen brothers, or David Lynch: Even if you don't like them personally the fact remains they have an audience, and they most certainly weren't made to order based on market forecasts. They represent the auteur theory in action, being the products of distinctive creative minds who impressed their equally distinctive visions upon them.
who? readers of a novel?
It was a Greatest Hits album of everything the stereotypical Boomer would consider great and memorable about their childhood and young adulthood.
What you call "exploiting a demographic" is more like "having a target audience instead of trying to be all things to all people". Sure, Boomers are a huge generation. But they were entering their 50's when Forrest Gump came out... and studios usually aim at 18-34 year olds because that's who spends the most time and disposable income at the movies.
What you call "Spielberging" is probably just a combination of cinematography, soundtrack, and accessibility over hipster snob appeal. In other words, good mass-market filmmaking.
Maybe this particular movie isn't your cup of tea, but there's nothing negative or cynical about this. "Cynical" would be hijacking HN threads to bring up urban walkability, UBI, random potshots at Baby Boomers, or any of the other standard hot buttons that typically guarantee upvotes and validation.
There is no "best" art (sorry, high art world), but there are absolutely choices you can make for marketing rather than artistic reasons.
I like Robert Zemeckis and Ron Howard films, but they're boring ideas, well-executed. And I have a sneaking suspicion that given their copious directorial output and uniformity of end product... some of those choices were made for monetary rather than creative reasons.
Nobody says "I want to make a mind-bending Dune film, get me Robert Zemeckis!"
That is fair. I believe it's also the case that when evaluating art, it's easy to conflate those reasons with one's own POV and demographic biases.
In other words: Forrest Gump is cynical pandering to Baby Boomers... but [this-year's-Marvel-superhero-movie] is an artistic gem, and The Dark Knight or Black Panther got robbed at the Oscars, etc.
"ars gratia artis" doesn't apply to Forrest Gump. It was manufactured out of marketing data. It was the 90s version of most recent nostalgia film "Ready Player One". I know it won a bunch of awards ( which I put no credence or value on ) but it is film saccharine. Forrest Gump is one of the forefathers of the current movie industry.
A filmmaker will flagrantly appeal to the cynicism of the producers in order to produce a film they believe in wholeheartedly. Forrest Gump is probably a great example of it.
Spielberg and Ron Howard definitely make middle-of-the-road appeal-to-everyone films.
But so did Shakespeare. Middlebrow and excellent are not mutually exclusive terms.
Oh, and Michael Bay is represented on the Wikipedia List of Film Auteurs page.[1]
Children > parents.
If you are a parent, don't you go to a child friendly restaurant that is reasonable good instead of one that is very good but not child-friendly.
Lots of "good movies" (using any definition you want) fail to find an audience. And even more mediocre movies make huge buckets of money. If your goal is to make money then just blindly making "good movies" is probably not your best strategy.
Box office returns dominate the return for a film, set the license price for long term post-theatrical deals, and decay with a very fast exponential from the opening weekend take. It is extremely rare in modern times for a film to turn a poor opening weekend into a high long term return. The film market simply doesn't work that way however nice it would be if it did.
I think most people can enjoy a really bad movie. Some t-and-a, explosions, a bit of CG - it's all good fun even all you can think about after the movie is how awful it was. But when really bad movies are all there is, they quickly lose their charm. Instead of being one off guilty pleasures, they become the movie industry. And that's about the point you stop watching movies at the theater.
I think this describes many people's transition away from movies. The point of this is that pumping out crap may be optimal in the short run. But in the longrun you end up with suboptimal returns, even when profit is your one and only motivation. Incidentally the same story is true and also being played out in various other industries.
Optimizing for next year's quarterlies is often not the same path to optimizing for earning the most money in the next decade.
That's one of the reasons why we have so many poor movies (the other reason is that making movies is hard). If your only consideration is making money, you'll produce whatever crap is cheapest to make that still makes reliable income. This is why e.g. consumer market is full of barely-working near-disposable products which existence is a pure waste of natural resources.
Seriously. There's no rule that says that the most profitable product is also the best for the user. In fact, it's almost always the opposite, and A/B-testing things to death leads to the former type of product.
Studios want what's "proven". It's why they will gladly feed in ML data to get shitty focus group type movies. They will then pour huge amounts of money in to try and get back even more.
"Good" is not a single-variable problem, you know? Sometimes you have a good script, but you don't have the perfect cast. Sometimes you have decent actors and all but almost no marketing budget so nobody hears about your movie. There's so many things that can go wrong even with the best intentions when trying to make a good movie that "just make a good movie" sounds really naive.
If it were that simple, every "good" director would make great movies one after the other. I struggle to find more than a few who could pull that off (and even then, that's highly debatable), among the the tens of thousands of directors who have been active in the past 50 years.
A drama like Shawshank Redemption (a favorite of mine) is not the type of movie I would have paid to go see multiple times in the theatre. Where as I saw The Matrix 6 times in the theatre back when it came out. Every friend who wanted to go see it I went with.
I'm professional film critic and I have a good idea of the data, because I analyze the feedback from festivals and industry experts, who decide what gets released and what not. From the Top 10 typically only 3 get selected.
The selection criteria has nothing to do with the quality of the movie and rarely the quality of the trailer, even if it's more important.
Most important are the names. Hollywood divides the Actors and Directors into A+, A, B+, B and C. Without an A name you'll have a hard time to get financing and proper distribution.
2nd most important criteria is the marketing power. How much money will be spent on marketing. If the distributor is not sure he'll try smaller tactics, like smaller releases, with the risk to earn much less. Most important is only the first weekend. Really good movies without names rarely exist, which can drive marketing by its own. Only these movies can survive a bad opening and raise slowly over the weeks to get profitable.
Hollywood knows exactly the triggers beforehand, that's why they can get away with so many formulaic and simple movies still bring profitable. There's not much need for ML, just counting the numbers and trusting the experts is enough. In this case, if a trailer works for which group is certainly helpful though. With trailers you get better feedback than from test screenings, but not much better feedback than from experts.
https://www.nytimes.com/1995/05/25/movies/gump-a-huge-hit-st...
Exactly, just don't write bugs.
The same way Michelangelo sculpted David, he just removed the marble not part of the statue.