Showing posts with label AI. Show all posts
Showing posts with label AI. Show all posts

Thursday, January 9, 2025

The online media landscape (and the looming drought for creatives)

From a young age, I always wanted to be a writer and/or an artist. Like most kids, I went through a variety of 'dream careers', but those two were always on the list, even if it was part time in addition to having a day job. I also always knew that my chances of 'making it big' in either of those careers were slim - as they rely almost as much on luck as they do on talent or skill, and who you know is often as important as what you know - but even until a few years ago, it seemed like a creative career was at least somewhat achievable, as long as I put in the time and effort to hone my skills and actually finish a project.

Unfortunately, the creative industry has changed a lot in the last few of years. Most of the following examples in this post are from an author's perspective, but many are just as applicable to artists in a lot of ways.

Once upon a time, if you were lucky enough to get your book published through a traditional publisher, they would not only print and distribute the book, they would also market it for you. The channels they advertised in obviously changed over the years - from paper catalogues to websites and finally to social media in the last decade or so - but it was still considered the publisher's job to sell your book. They would keep a greater percentage of the profit from sales of your book, but they were also the ones doing the work and taking the financial risk, so that was at least somewhat reasonable.

But at some point in the last 5-10 years, that responsibility seems to have been shifted onto the author. Unless you're one of a very few, very big names (eg. Stephen King or N. K. Jemisin big), your publisher is unlike to do much advertising on your behalf aside from the odd Tweet, which is likely to be as effective as shouting into a well. Brian Kirby did a thorough comparison of engagement that various publishers were getting on Twitter versus Bluesky for promotional posts they made. They found that even with only about a fraction of the follower numbers, the number of likes and ReTweets the publishers got on Bluesky were significantly higher than they were getting on the same posts on Twitter; in one case, the only engagement a Tweet got was from the author of the book being promoted.

Some publishers will apparently take it a step further and blame the author for not promoting their book if it happens to have poor sales. In other words, they expect you to do their job for them and pay them for the privilege of doing it. Not only that, but some publishers won't even touch you if you don't already have a huge social media following, which makes being a writer feel a bit like being a job seeker who can't get a job because they don't have experience and they can't get experience because no one will give them a job.

So it seems like these days if you want to make it as a writer (or an artist), you need to have a good social media presence. You need to be a promoter and advertiser in addition to being a writer (and probably in addition to your day job as well, because the vast majority of us can't afford to write full-time) and have the skills to market your product to your intended readers. The problem with that is that the same social media sites you have to rely on to find your audience seem to be hell bent on alienating as many users as possible.

CultureCrave posted on Bluesky (via Financial Times, though going direct to the link seems to hit a paywall) that Meta plans to fill Facebook and Instagram with AI generated users*, which to me seems akin to a sports centre manager announcing that he's going to start shitting in the swimming pool as if that's a good thing. Meta claimed they were doing this to "make their apps more engaging" and attract more users to the platform, even though filling the apps with bots will have the opposite effect. It's bad enough seeing the deluge of AI-generated slop posted by accounts run by humans (not just because of the poor quality but because a lot of it is just downright wrong) added to the existing slurry of advertising spam and constant harassment to follow pages or join groups I'm not even interested in, but if the accounts themselves aren't even real, what is the point of interacting with them? What is the point of staying on that platform? While many users will stay out of apathy or because these platforms are the easiest or only way for them to stay in touch with certain friends or family members, a lot will simply leave. Some might end up on the shores of some other social media platform (where it's going to be difficult to start building up connections again, especially if your friends and followers scatter to multiple different platforms), while others will just vanish outright; I left Twitter** along with many other users as part of last year's mass exodus, and although I did find some of my Twitter friends elsewhere, many of them have apparently disappeared into the ether.

I've already noticed the effect of this social media attrition even on my own small pages. I run dedicated profiles for my art on Facebook and Instagram (and on Twitter, before I nuked all my posts), and once upon a time, when I'd post a new art supply review or demonstration, I'd get at least 10-15 likes on each (not much in the grand scheme of things, of course, but at least I could see that my work had some reach). Now I'm lucky to get 2-3 likes, and those are usually from the same handful of friends who like all my posts. If my already-established public pages are facing such a significant drop in views, it's going to be so much harder for a creative person setting up a new page and trying to build up an audience from scratch. And at this stage, there aren't really any viable alternatives; sure, you can set up your own website (or even a blog), but how do you get people to find it?

As a creative person it's actually really goddamn depressing knowing that even if I manage to finish writing my novel, in order to get anywhere, I'll be forced to do my own marketing and advertising because the people whose job that actually is (ie. publishers) don't do it anymore, which means having to rely on social media that is becoming more and more unusable because the companies that own these sites are just pumping them full of spambots and AI trash (like this and this) and letting abuse and hateful rhetoric against minorities flourish***, thereby driving the actual audience away.

Maybe one day things improve for artists and writers (and other creatives), but I fear things will keep getting worse before they get better. I can't help but wonder how many awesome books and how many beautiful drawings and paintings we won't get to see because they got drowned out by the sea of ChatGPT and MidJourney trash or because publishers wouldn't even give them a chance. I don't know a single sensible person who actually welcomes all this AI rubbish in place of human-made art and stories (as I've posted before, it's not just poisoning creativity but also people's ability to think critically), so I just hope there's enough of a pushback against it that we as a society can prioritise real imagination. As for traditional publisher's failing to do their jobs, I know there are self-publishing options and the ability to create eBooks, but as with having your own website, how do you market your book to an audience that isn't there?

*Sure, Meta has since deleted these AI-generated profiles in response to the overwhelming backlash against them, but if you believe they won't try something similar as soon as they get half a chance, then I have a bridge to sell you...

**I still have an account there to maintain my username, but I deleted all of my posts and removed the app from my phone and I no longer open the site.

***[content warning for this footnote; threat of sexual assault] Let's be honest, they've always done this last one - Twitter once told me that someone threatening to rape me with a broken bottle for saying I didn't care for Steven Moffat as a Doctor Who showrunner "didn't violate their community standards", and Facebook have repeatedly refused to remove pages encouraging violence against women (though they did put a year-long warning on my account for "threats of violence" when I joked to a friend about "nuking" my account!) - but now many social media companies aren't even pretending to care about the safety of their users anymore, with Meta now explicitly allowing abuse and harassment of women and queer people and removing their fact-checking system, Twitter being taken over by a giant attention-seeking toddler who supports a convicted rapist and felon and allows misinformation to run rampant and BlueSky refusing to remove a user known for harassing trans people (as well as other toxic behaviour).

Thursday, September 5, 2024

NaNoWriM-Oh no: Why AI is bad for creativity (and for everything else)

At the start of September, NaNoWriMo made clear their position on the use of AI in writing, and though it seems they've edited the wording slightly since the furore around their statement began, they still seem oblivious to why people are so opposed to their views.

Before I launch into my own anti-NaNoWriMo rant, allow me to provide some background information:

I am an academic, and a teacher at a university. I teach a variety of units, but there is one unit in particular that I have taught for more than a decade, both as a tutor and as a lecturer (I have also developed and run an online version of the unit for my university's online-only course branch). So I think it's fair to say I have the experience to recognise trends and patterns in the student cohorts in terms of their achievements.

Based on what I have seen, I firmly believe that the introduction of AI tools such as ChatGPT will set back humanity's development and advancement by a decade or two. When we set a task in class for students, instead of engaging with one another and discussing the concepts, they open up ChatGPT and paste the tutorial instructions into it, and when we ask them to share the answers, they just regurgitate the slush ChatGPT spat out. The problem is that while ChatGPT is very good at producing content that sounds reasonable, it is not good at nuanced thinking or self-reflection or considering scenarios that might fall outside the norm. No matter how many times we pull students up on their incomplete or in many cases incorrect or inappropriate answers that they got from ChatGPT, they would still rather rely on a flawed system than try to apply the content or principles. In other words, they seem to have lost the ability to think for themselves (and now there's research that backs this up). This isn't just a problem in education but has wider implications for society as a whole.

This is particularly evident when it comes time for me (and my colleagues) to mark assignments. Because many of the students did not actually complete the activities in class, they didn't learn the skills required to complete the assignment tasks properly, which means that instead of meaningful deliverables and insightful analysis of what they have done, we end up with pages and pages of word vomit that use a lot of big and fancy words (if I took a shot every time I saw the word "meticulous" in an assignment that was very clearly not done meticulously, I would have liver damage) but don't actually say anything of any substance or value. It's been a few semesters since ChatGPT became widely accessible and when I say that the average student marks have dropped by a full grade since that point, I am not exaggerating.

So when I saw that NaNoWriMo had come out in support of people using AI in their 'writing', I wasn't particularly surprised because of how problematic they've been in recent years (and I'd already decided I wasn't going to participate again because of the appalling way they handled those incidents), but I was disgusted.

I've linked to their statement at the top of this post, but I want to focus on three sections in particular that stood out to me.

"NaNoWriMo does not explicitly support any specific approach to writing, nor does it explicitly condemn any approach, including the use of AI... We fulfill our mission by supporting the humans doing the writing."

Aside from how pathetic and wishy-washy this comment is, it's contradictory and also demonstrates a lack of understanding about what AI actually is, and how it works. Content generated by ChatGPT and similar AI tools doesn't just magically come from nowhere. It is built on stolen work. Actual artists and writers created this original content, and ChatGPT just chews it up and spits it out without providing any acknowledgement or compensation to the human beings without whose work it couldn't exist. You can't claim to "support the humans doing the writing" when you allow or encourage the use of a tool that does the exact opposite of supporting actual human creators.

On a side note: Writers who complain on social media about your writing being fed into AI but then include AI-generated 'art' in your social media posts? You are part of the problem. You cannot claim to be upset about your work being stolen when you are turning around and doing the exact same thing to other creatives.

"We believe that to categorically condemn AI would be to ignore classist and ableist issues surrounding the use of the technology, and that questions around the use of AI tie to questions around privilege."

This bit actually made me snort. Implying that disabled or poor people can't write without AI to help them is far more condescending and ableist/classist than criticising the use of Artificial Idiocy ever could be. In fact, the poor or disabled people NaNoWriMo claims to want to support in their ridiculous statement are also among the most likely to be disadvantaged by the existence of these AI 'tools', because they're far less likely to have the resources (time, energy, money) to fight back when their work is stolen and passed off as the magnum opus of some pretentious wanker who thinks they're going to be the next Hemingway just because they mashed a few buttons in ChatGPT. Frankly, if you can't write stories without using a machine to steal bits of other people's stories for you, that's not a case of "ableism" or "classism". That just means you're not fit to be a writer.

"It's healthy for writers to be curious about what's new and forthcoming, and what might impact their career space or their pursuit of the craft."

At this point, writers who genuinely care about the craft have a pretty solid understanding of how AI might impact their career space. Spoiler: It's not good. I follow many artists and writers on various social media sites, and I have not seen a single positive comment about AI from any of them. It's not just that AI steals the content from the original creators without paying them. As with my students, many people would apparently rather have something crap but fast and easy than put in time and effort or pay for something that is actually worthwhile. AI 'art' is the fast food equivalent of creativity: Sure, you can have it quickly, but it has no value and you'd regret consuming it if you actually thought about it for more than a minute or two. The increase in people turning to AI to pretend to make things for them means the people who actually make the art or write the stories you love aren't getting paid, and if they're not getting paid, the industry is no longer sustainable for them, so they will just stop creating; that means less new content for fans.

And it's not just the financial impact on the creators. True creativity is what differentiates us from machines. The need to make something that evokes feelings and provokes reflection is something that only humans have. The desire to grow and improve and become good at something is what lays out the pathway for a kid scribbling away in their notebook to practice and learn and eventually make something that only they could have made, because it has come from their experiences and their thoughts and their emotions and is, in some way, a window into their soul. If we take the soul out of art, what's the point?

Anyway, it's disappointing that an organisation that used to be a fun and engaging way for writers to communicate with one another has turned into *gestures vaguely at the festering corpse of NaNoWriMo's integrity* whatever this is, but I think it is also now pretty clear that NaNoWriMo is no longer worth your time or money.

I just deleted my NaNoWriMo account (which I should have done years ago but just never got around to it), and I suggest you do as well.

EDIT: Some arguments I frequently see from people trying to justify the use of ChatGPT etc are:

  • AI is going to take people's jobs so they should just accept it and work out how to co-exist with AI instead of being in denial and fear.
  • People already repurpose other people's work and call it inspiration.
My response to these arguments is:

The thing about people taking other people's work as inspiration is that even that reimagining of an idea is still based on that person's own experiences and history, and things that resonate with them. Whereas AI just takes everything. It's like making a soup out of every single ingredient in your cupboard instead of just choosing the few ingredients that actually work well together.

And as far as taking jobs, why are we automating creative jobs when there are multitudes of people willing and able to create good art/stories etc instead of automating the boring and tedious parts of jobs (or life in general) that no one wants to do (which would actually give people more time to enjoy life and do the things that matter)? The people coming up with this AI tech think they're entitled to other people's creative output and that it's okay to just take it because they don't see it as having any value, which I guess is why they consider the generic crap it spits out as "good enough".

While AI might theoretically improve to the point it becomes good at doing things, the fact it's not there yet but people are using it anyway is why I feel it's not a good thing for advancement. AI doesn't seem picky about what it uses, so the more crap AI puts out, the more crap AI will consume when it's trying to generate new content. Too many people seem happy to just accept the junk ChatGPT spits out so they're just going to start becoming reliant on it instead of actually thinking critically about things and figuring things out for themselves and actually finding creative solutions for things, which is basically how humanity got to where we are now.

On a side note, even if AI somehow did magically become useful, is the cost actually worth it?