AI - the poisoned well
Ex-Google CEO Eric Schmidt got booed at a graduation this week.
If you haven't seen it, you should. It's awkward.
The booing was not really for just for him. It was for the entire class of people who have spent three years telling us their tools are simultaneously the most powerful in human history, and are going to take our jobs and kill us all.
Glorious upside if you are running one of these rapidly scaling companies, or (more to the point) holding shares in them. For the young people in that crowd, who are told daily that their expensive degree is worthless and no one will hire them, the joke is rather less funny.
This is the most acute have-vs-have-not divide of our age. Disenfranchisement on a genuinely existential level. Brought to you, ironically, by many of the same people who already have more money than most countries.
For the record, the contents of what Schmidt said were not wrong. Feelings about AI have just become so negative that it is hard to even mention it. The boosters have poisoned their own well.
Do I think AI is useful? Yes, I do.
Do I think it is as valuable as the boosters say it is? Erm…
A 5% flier
I get asked what I think about AI production and how it is affecting our business, all the time. The honest answer is yes, it is affecting us. Not in the way you would think.
It is not that every business is suddenly making stunning GenAI videos with very little effort. They are not.
It is that the execs at many of the businesses we work with think they are. And that belief, on its own, is doing more damage to good content than the technology has.
Last summer, an MIT study found that 95% of enterprise generative AI pilots produced ZERO measurable impact on the P&L.
None. Zilch. Nada. (Forgive the emoji but seriously 🤯)
Now, the models have improved a bit since then, but still. The boosters are selling you a 5% chance of any positive return at all, and have convinced your CEO/CFO it is the strategy. It is the sleight of hand of our age.
Our clients now hesitate to sign off on a $150k film because the hype suggests a $500 AI solution may be out there. Press a button, get a great output. Much as some of the publishers on this benighted platform would have you believe.
At a time when everyone is already in fear of their jobs, that is quite some doubt.
The fact that the project in question is a week long documentary shot on location in Sub-Saharan Africa does not seem to dampen the concern. The hype has reached the point where physics is optional.
Filmmaking - a human endeavour
As I keep saying when asked, if you want to make a really good AI film, you need a really good filmmaker. Or filmmakers...
The best AI films follow the same process as traditional production. A (human) scriptwriter. A (human) producer. A (human) director writing prompts to a (synthetic) generator. A (human) editor. A (human) animator cleaning it up. A (human) sound designer to make it gleam.
Sometimes one human plays more than one role, but you get the point. That is a whole lot of flesh and blood.
The crane and the mason
I have made this point here before. AI is a faster crane. A hell of a crane. But cranes alone do not build cathedrals. Stonemasons do. Without them, all you get is a bigger pile of rocks. And a bill.
In the right hands, you can create things that would never have been possible within the budget. We're making lots of AI films in Asia, where the progenitors haven't been telling everyone one that the platforms that have made them stupendously rich are going to kill everyone (after taking our jobs).
So the public are a lot more positive about the use of AI.
Yes, these films take a lot of prompts (which are about to get a lot more expensive). Yes, they look a bit AI. Yes, the audience may discount them the moment they realise they have been duped.
But pressing a button, they are not.
Good filmmakers make good films
A friend of mine who runs a small business took an entire week off to make a film for his company. The result? I probably should not type the exact language he used here.
And he is not the only one. I know loads of people who have been lured into this hype loop and have nothing to show for it except a chunk out of their life they will not get back.
Can AI give you decent filmic results? Yes. But you will need a decent filmmaker and a process to do it.
So perhaps those grads getting told their degree is worthless do not, in fact, have as much to worry about as the hype would like them to believe.
The button does not exist. The boosters are still cashing in on the fear that it might.
Have a great week in our real world. So amazing you can smell, taste, and feel it.
See you next time,
Oh, and if you have shipped a decent film by just pressing a button, I would genuinely like to hear about it.
I'll wait.
.svg)
