Leon Furze

In 2024 I posted at least two articles a week on this blog, sometimes tipping over to three or four if something particularly interesting happened, or yet another new release, AI application or feature. And then towards the end of the year, it came time for me to shift gears and focus on writing my PhD thesis. The more I concentrated on the PhD thesis, the less interest I had in the blog and related writing like social media posts.

This is nothing unusual. After all the energy poured into the blog in 2024, I think my brain just needed to switch off from that sort of writing for a little while. Other factors were at play. As I’ve written about before, I draft a lot of my blog posts, including this one, verbally while I’m out for a run or a walk on the farm where I live. During the summer months in Australia, I run outdoors less often because of the heat and the frankly terrifying number of tiger snakes that I have to navigate on the road.

I also found it hard to compartmentalise my thoughts on AI in the way necessary for short form writing like the blog. My PhD involves a lot of discourse analysis, which in this case meant getting up close and personal with other people’s thoughts, their concerns and excitement, use or rejection of the technology. That research is a place where my opinions on AI, frankly, are just not that important.

Dividing Lines

The public discourse around AI also feels as though it’s become even more divisive since Christmas 2024. Perhaps it’s the political climate in the US and the way technology companies seem to be falling over themselves to toe the harmful and frankly childish anti-woke party line. Between Zuckerberg’s “masculine energy“, Open AI’s relaxed guardrails, and Elon Musk’s… well, everything… there’s not a lot to like about the major AI developers at the moment. Open AI hammered a fairly hefty nail into the coffin of respect when even the CEO jumped on the disturbing Studio Ghibli bandwagon after the release of their updated image generator.

On social media, where I’ve also been posting much less frequently this year, you have to be pretty thick skinned at the moment whether you’re posting for or against AI. On the occasions I’ve posted something critical, such as calling out Open AI’s changed policies with regards to offensive images, I’ve been attacked by the “pro freedom of speech” crowd. I thought they’d all drifted off to little private island made up of x.com accounts, but apparently they’re lurking in wait on LinkedIn and even Blue Sky.

If, on the other hand, you post something even vaguely positive about AI, such as a fairly innocuous post stating that Chat GPT’s new image generator can produce .png files with transparencies, it gets swarmed by disgruntled artists. I side infinitely more with the artists than the spurious freedom of speech, anti-guardrails crowd. But it’s difficult to drum up the willpower to post anything either useful or critical when you have to spend an hour manually replying to or deleting antagonistic comments after every single post.

Of course, this is all just the privileged whinging of somebody who has drummed up enough followers on LinkedIn for the algorithm to care in the first place. I can’t really spend the whole of 2024 building up an audience and then complain in 2025 when they respond with their opinions.

Running Back to Writing

So, on this uncharacteristically warm autumn morning, I’ve decided to shamble (physically and metaphorically) through this post to see if I can get my thoughts back in order.

Whenever I lose the will to run or to write, I reread a short non-fiction book by Haruki Murakami, What I Talk About When I Talk About Running. I’ve read the book so many times that the pages are falling out, which is perhaps an indication of how many times over the past 10 years I’ve stopped and started both activities.

In the book, Murakami pulls together a collection of disparate essays and scraps of writing; memoirs about his thoughts as someone who runs daily, completed at least one marathon a year for a decent chunk of his life, and of course, has written dozens of fiction and non-fiction books.

It only takes a few hours to read and the same few quotes jump out at me every time. This has probably been the most important quote for me over the years of any book. It pushed me through chunks of dissertation writing, journal articles, and the 2024 blog post spree, leaning on extra work, posting on social media and struggling back from 2K to 4K to 10K.

Once you set the pace, the rest will follow. The problem is getting the flywheel to spin at a set speed – and to get to that point takes as much concentration and effort as you can manage.

Haruki Murakami – What I talk About When I Talk About Running

This post is one of those first pushes against the flywheel for 2025.

Perhaps oddly, for someone who’s made a living of writing, Murakami also talks about the toxicity of the writing process. How writing is, for him, an inherently unhealthy process and something which needs to be counterbalanced by the physical activity of running.

Basically I agree with the view that writing novels is an unhealthy type of work. When we set off to write a novel, when we use writing to create a story, like it or not a kind of toxin that lies deep down in all humanity rises to the surface. All writers have to come face to face with this toxin and, aware of the danger involved, discover a way to deal with it, because otherwise no creative activity in the real sense can take place.

Haruki Murakami – What I talk About When I Talk About Running

Writing is obviously sedentary for the most part. Although I write these blog posts on foot, I find I can’t do the same with fiction or something as involved as my thesis, where I need to have all of my notes to hand. Even the final work of editing and posting these verbal drafts happens at a desk.

But Murakami is also talking about something more than just the physical exertion, or lack of, involved in writing. He’s saying writing is a process where you store up some of the toxicity and negativity and allow it to seep through your pores onto the page. Murakami talks about fugu fish, which have the “tastiest part near the poison”, but there are plenty of animals that absorb harmful chemicals from their environment and then excrete those compounds as a defence mechanism. Wikipedia tells me its called sequestration.

I agree with Murakami’s opinions on writing, and I think that like a poison dart frog, many writers secrete the various harms they’ve collected as a form of defence. I’m obviously not the first person to acknowledge the therapeutic benefits of writing, but I quite like the idea of writing being an activity which is fundamentally risky.

AI and the Risk of Writing

And that brings me back around to artificial intelligence, because what generative artificial intelligence does is neutralise the toxicity of writing. It removes the risk, the danger of taking your personal thoughts and expression, your ideas, what you really think and how you really want to say it, and running it through the artificial grinder of a large language model adds a layer of abstraction between the writer and the final product.

And sometimes I’m okay with that. I’ve contributed to academic journals where I’ve used generative artificial intelligence to edit and clarify paragraphs and help synthesise disparate ideas from literature, or to take my personal opinion, restructure and rework it until it’s something more objective, more rational.

If I’m writing an email, nine times out of 10, I want it to be clear and direct, not passionate, meandering and filled with ambiguity. If I’m filling out an application form, and I have to respond to specific criteria, I want to make sure I’m actually responding to that criteria, not just the story I’m telling myself about any perceived hidden meaning. Government forms don’t necessarily require me to read between the lines.

But elsewhere, on the blog, my thesis, and on the rare occasions now that I still write fiction, I feel like I want the poison to seep out a little. There needs to be some vulnerability. Perhaps frustration about being attacked on social media, irrespective of which side you choose. Maybe a sort of apologetic, humblebragging about the fact that I’m out here going for a run. Like Murakami writes in Running: “I think that one more condition for being a gentleman would be keeping quiet about what you do to stay healthy.”

Oops.

I want to see that discomfort represented in the finished piece of writing, not smoothed over by a predictive algorithm.

It’s taken me the first four months of 2025 to have this real first push on the flywheel. I’m about to start the slightly daunting task of updating my Teaching AI Ethics collection for 2025, revising the nine original areas and adding everything that’s happened for good and for bad in the past two years.

I needed something to kick start me back into the writing habit and to shift gears again, away from the slow, thoughtful, desk-bound, thesis writing to the personal, reflective and decidedly more energetic blogging. This post is self-indulgent, selfish, and probably a bit toxic. Exactly the kind of writing that AI is hopeless at.

Want to learn more about GenAI professional development and advisory services, or just have questions or comments? Get in touch: