Welcome to the AI For Humans newsletter!

It’s nearly Christmas and lots of people will be spending time with their families this week.

Families love to dig in on current topics and since most of us learned to avoid politics a few years ago, it’s possible the chatter around the dinner table might be about AI.

In the last month, I’ve been tracking the rise of a number of different anti-AI narratives that seem to be crossing over to a very mainstream audience.

From gamers hating on AI use-cases in game development (the Larian Studios’ ‘controversy’) to the re-rise of the ‘AI will use all the water’ argument to the ever present ‘AI will kill us all so why are we even pursuing it’.

So today, I thought I’d quickly give you a few positive narratives and maybe leave your family feeling better about the state of the AI world, specifically around creative work.

The Larian Studios situation is, to me, a perfect encapsulation of how heated the idea of ‘AI anything’ has gotten for some people.

A game studio known for their incredible human creativity and voice acting (please play Baldur’s Gate 3, it’s a masterpiece) mentioned that they use AI tools in the very earliest stages of brainstorming and conceptualizing their games, got dragged hard by the gaming community for even hinting that AI use in gamedev could be ok.

What’s underlying this argument, at least in my opinion, is the idea that AIs are coming for creative work (and income) and that people who love games (and, by extension, movies, books & everything else in the same bucket) here are defending the humans making the stuff they love.

This isn’t a terrible instinct! We should make sure that creative people get to keep making good stuff! And there’s been a ton of layoffs in the gaming and entertainment spaces already.

I’d argue, however, the economics of corporations and the shifting buying habits of these same gamers is what has shaken the business rather than AI.

Love this newsletter? Forward it to one curious friend. They can join in one click.

But let’s assume that those points won’t make a dent (they’re harder to prove unless you want to bring a full stack of links).

So what can we say back to these people when they ‘I will never play a game with AI!!’ or ‘they’re all scammers and they’re trying to rip everyone off’?

Well, first and foremost, I’d immediately give them the point that the initital training of these models from the large companies (Google, OpenAI, Meta etc) is controversial and legally still an open question mark. Texts and images were scanned and injested without direct approvals and there are on-going lawsuits around these issues.

Maybe you can give them a little snippet about how Anthropic agreed to pay 1.5 billion dollars to settle a lawsuit with a authors of books it used for training.

But that’s not the main point here.

Start by opening the door to the idea that these large AI companies themselves are not making games or AI videos or anything else. OpenAI, Google or Mata are providing tools that others are using to make these things but they do not make them on their own.

Even in a future AGI world, even then, these tools will not be making creative works entirely on their own. Why would they?

What interest does an AI system have to make a creative work? It’s not going to watch or play or read it. Maybe, maybe, in some advanced world a Super intelligent AI would see creative work as a way to pay for it’s ongoing compute costs but that’s a LONG ways off.

Even in that advanced world, the human is the one who’s going to ask for something very specific to play or watch.

Or, more realistically, a human might have an idea in their heads, tinker with these tools for a while, finally create something that gets close to what they wanted and then what?

Most likely they’ll release it into the world for others to try. After all, they spent a ton of their own creative energy on it and they’d like other people to play/read/watch it.

Humans make things for other humans.

Even the worst types of AI hypebeasts (those espousing ‘automated-content farms’) are the ones creating the systems. Even that is human ingenuity and creativity at work.

Once you get across the idea that AI is not some specific boogeyman (boogeyalogrithm?) responsible for all the terrible stuff in the creative world and that we are the ones powering stuff for us, the next big step to me is getting people to think about how anyone makes anything.

As someone who’s made a fair amount of stuff (from TV shows to start-ups to this very newsletter), I can tell you clearly that the act of making something isn’t the idea and it isn’t the final product (though that is what you see), it’s all the little steps and decisions along the way that I’ve made to make the thing mine.

And, as someone who’s made a fair amount of stuff with AI, I can also tell you that, in general, AI is terrible at making things feel like mine.

When I first started up this newsletter in earnest again, I tried to have AI write this with me (not for me!) and it just made it remarkably worse.

But it does help me tease out some of the ideas I might want to write about and helps me think through some of the jumble of thoughts in my brain. I then take some of that help, write a few more things (myself mind you) and eventually get this to a place where it’s somewhat readable.

This is all a long way of saying that all AI content is actually human content but good AI content is even more human content because you can feel the human work from within it.

In my opinion, the fear that a lot of people have around AI work has less to do with the process of it all and more about whether or not their art or work will be recognized as good or something people love.

My hope (and I think the big positive thought to take away from this) is that while AI may let us make more stuff that we as humans will continue to recognize work that others humans do and reward them in kind.

And while that may not win the argument at the dinner table, it might slightly crack open a few closed doors and get new people trying to make something themselves.

Have a wonderful holidays & we’ll see you in January!

-Gavin

In this week’s AI For Humans: ChatGPT’s new Image Model & More!👇

Did we mention it’s Christmas this week?

Well, OpenAI has a little treat for you within ChatGPT itself.

Type a present emoji (or cut-n-paste this: 🎁) into the text box within ChatGPT and you’ll get a new little app that creates a Sora video based on a picture you upload.

Merry Christmas y’all! Hope you have a nice time during the holidays!!