The rise of AI in creative industries

To begin I should probably point out how incredibly biased I am, I am a software developer and a screen/stage writer. Both industries are currently under threat from the various AI solutions. I have also worked on AI development, albeit early in my career.

There has been a recent flurry of activity in the use of AI solutions, most of these are being driven by various University programs – matured through partnership with corporate sponsors.

This information isn’t just a throw-away comment, it helps in understanding why projects like ChatGPTx seem to be all over the internet. It’s being pushed to help raise its importance, acceptance and, eventually, it’s value.

Now, ChatGPT and other similar technologies aren’t bad in their own right. They are interesting examples of expert systems and the various training databases that are being built.

The real problem comes in how they are being sold. The advertising for this technology positions it as an actual Artificial Intelligence, something that can create answers, can solve problems and create art.

None of these things are actually true. In reality, these systems aren’t especially more intelligent than the systems that have come before, the real difference is in the technology and data surrounding them. Better and cheaper storage, faster processors and the vast database that is the internet. These are assets that simply weren’t available to researchers in the past, and it makes a vast difference.

Let’s start with one of the least controversial aspects of current AI systems. There are a bunch of blogs, social media posts and videos showing people interacting with the various AI text bots, asking them questions, asking for opinions and asking for information.

Used this way you can be fooled into believing you’re chatting to an intelligent system. The vast troves of data and improvements in natural language parsing make for an interesting experience. But you only have to go slightly off-piste to see the cracks start to appear. Start making subtly non-sensical requests and it will get very confused. In situations where human intelligence and experience would allow some kind of response, even if that response was ‘what are you talking about?’ the software doesn’t have anything, it’s database of human communication doesn’t have a relevant base for it to work from.

Possibly more dangerously, it has no way of tempering the information it has, the internet is a fascinating place, but more than a little of its information is incorrect, not always accidentally.

And this all leads to the main point of this blog, the current crop of ‘AI’ software can’t create anything.

People will argue about that, they’ll show the various AI-generated images, the scripts generated and the music composed. But none of that was created, it was generated using work pulled from the internet. Art from various artists, scripts from scriptwriters, and music from composers all mashed together using a formula to make the appearance of something new.

You may not think this is a problem, you might point to parody songs or covers. You might talk about how TV and Film are already derivative and how art is in the eye of the beholder. But all of these arguments rely on one important fact, Musicians, Writers and Artists exist and are creating new works.

This sounds like an overreaction, but this is a very real problem. Let’s concentrate on the part of the industry that I know best, writing. Being a TV/Film writer is already a hard job. The pay isn’t great and it’s really hard to break into the industry. As I write this the Writers Guild of America is striking, they are doing so because the situation is getting so bad that many writers can no longer write as their only job. At that point, writing can no longer be considered a career.

More importantly, the idea of using AI to replace writers has been raised by the studios, some even going so far as to produce scripts for shows based on an AI scraping scripts from previous episodes.

Hopefully, the issue here is obvious. If an AI, which can’t create new ideas, is writing a script based on previous episodes then you’re just going to get episodes that are rejigs or past episodes.

Again, many will argue this is already the case for much of the industry, but that is the fault of the studios, I can guarantee that every commissioned writer has something new they want to sell, but studios want to stay to a pattern, new programs scare them.

Now imagine the situation we have now, but one where there is NO chance of something new being created, because it’s all being made by AIs scraping existing scripts. That’s a very real possibility and absolutely one you should be concerned about.

AI, as it currently stands, is a threat to every creative industry and more needs to be done to control it. We’re already seeing spam being generated – allowing scam artists to send out huge amounts of fake comments and emails that look slightly more realistic than previous ones. We’re already seeing people taking credit for art that is made up of other people’s work and we’re already seeing studios using the threat of AI as a way to strong-arm writers to accept terrible conditions.

I saw a post on Mastodon once that stated that the people who were against AI were the same tech-bros who were for NFTs etc. This made me laugh because it’s all the same people. The same people pushing rug-pull blockchain ‘opportunities’ are the people backing these AI tech companies, because it’s all a similar scam.

The Retrospective: Wedding

Because i’m the height of romance and an agile freak I decided to write a retrospective on mine and Erica’s recent project … I mean wedding. I joke, of course, but given i’m an Engineering manager and Erica is a director it was always going to be run like a project and, I think, we benefitted from that way of thinking.

In the true spirit of an Agile retrospective, here’s the do more of, do less of and keep the same breakdown.

Do more of

Pick your venue carefully, not just based on cost and the actual facilities but also on how well you get on with the people running it.

Frankly, we struck gold here with The Wellington Arms in Hampshire, because it’s a triple threat in this area, it’s cost effective, attractive and the people involved were super helpful and friendly.

Be different.

We decided not to follow many of the conventions weddings normally hold to. Some of these were because of particular difficulties. For instance, we couldn’t have normal flower girls because the venue doesn’t like having things strewn everywhere, so we changed to bubblegun girls with jingly waving wands. I wanted my two closest friends involved so eschewed the best man idea and just had two witnesses. A friend who is gifted at photography offered to handle the photo taking. This meant we got photos that felt more personal, because the photographer knows us.

DIY.

Because of cost reasons we ended up making quite a few things ourselves, one of those I’ll mention in a different card, but we made Pacman Ghost shaped candle holders for the table centrepieces and they turned out great. We also designed all the documents (menu cards, order of service, notices etc) ourselves and used the opportunity to create a sort of wedding brand.

Created our own rings

This wasn’t originally about cost, but it ended up saving us a lot of money. I designed both my own and Erica’s rings. For my ring I just 3D printed and treated it, because my arthritis means I’d probably end up having to cut it off or remove it anyway. For Erica’s I used a 3D printing service to get it made in solid silver. Not only does she have an entirely unique design but it was also vastly cheaper than buying something of the same quality.

Do less of

Don’t skimp on the cost of the cake, or do it entirely yourself

We paid over £200 for our cake, which I felt was quite a lot of cash, of course it’s not. This, unfortunately, showed. The maker didn’t get the design correct and the actual cake didn’t taste any better than something we could have picked up from a supermarket. Fortunately, for us, the event company we used to get a few of the things we couldn’t make (outdoor games for the kids, flower arch etc) also had a baker working with them who, with the help of a friend and the chef at the Wellington Arms, not only fixed the design but seem to have added fruit and stuff to the cake we had on the day to make it better – they also did that for no charge. Needless to say, we also recommend BWH Events.

Don’t assume everyone understands technology like you do

I left some aspects of the technical side of things to people at the venue, and whilst they were amazing in many ways, they didn’t quite manage to get the bubble guns working in time. I should have handled that myself.

Don’t skimp on a videographer

To keep costs down, and because we happen to run a production company, we decided to do the video of the wedding ourselves.  This proved to be a mistake with the eventual video not being as good as it could have been.

It was like trying to direct a film that you were also main camera and one of the main actors for.

On the day, don’t try to be a hero

It’s fine trying to manage this stuff yourself, but you have to remember to let other people help you more. We did and our fantastic friends stepped up, but we should have let them do more.

Keep the same

Small service with later reception

This was something dictated by costs and the decision by the government to hold a coronation the weekend after our wedding. The smaller service was fun because it meant we were able to concentrate more effort on each guest, meaning we were able to spend time with everyone. I miss the fact I didn’t have my siblings there, but I didn’t feel it would be fair to invite any of them when I couldn’t invite all of them.

The later reception (a week later in fact) makes it feel like our wedding is on-going and culimating a week later with even more people. (at the time of writing this the reception is still yet to happen, but i’m hopeful it’ll go well).

Whilst it worked out well, I did miss having more of my family (and their families) there.