Not an AI - Statement of legitimacy

When I started out on the exciting and difficult task of being a novelist, I knew I was going to be spending most of my time fighting an uphill battle and that things would take a lot of work, a lot of time and that the road ahead was fraught with many unforeseeable challenges.

I approached the task by drawing on my experience as an analyst and as a consultant. Long story short, the first step you take when founding a business or entering a new market, is to become familiar with the target market's historical trends. You (1) check to see how things have been going for the last few years, (2) you see what other people did right and what they did wrong, and, lastly, (3) you try to figure out where the market is going.

I spent a great deal of time working on an answer to the first two questions and I also thought I had an answer to the third question.

I deduced that my market is not all of media or entertainment, since reading is an activity that takes place almost entirely within the mind of the reader (you don't 'see' a book like you do a movie or a video and you don't 'listen' to a story, as you would to a song). Hence, I realized that I was not competing with cinema or the music industry since those are inherently different experiences. A film is something that you observe on a screen in front of you; a book is something that unfolds inside of you. The words on the page in front of you are really just guidelines for a story that is actually happening inside your own head.

So, with this knowledge, I set about 'selling' that type of product and I focused a lot on studying how other indie authors got their first bit of traction. I figured out what my marketplaces should be (you only really 'need' Amazon and its affiliates), what my niche was (it's a psychological science fiction book that presents itself as the story behind the birth of an artificial afterlife, but is actually a story about inter-generational trauma) and what my stages of growth were (obtain good reviews, run ads, build an organic fanbase, obtain blurbs from established authors, partner with book influencers, appear on recommendation lists, invest in a professional editor, invest in an audiobook, invest in translations, etc.)

All of these things are hard enough as it is, though I must confess that there is another element that I did not foresee (at least, not accurately).

Artificial intelligence.


One of my favorite courses in university was called 'Science, Media and Society' and it taught me how to objectively approach every single news story that came my way, particularly when it pertained to new technologies. Scientists and engineers are inherently bad at explaining what they've just achieved to non-scientists and non-engineers. The media tries really hard to understand what the scientists have just come up with and then subsequently explain it to the public. And the public, well... they're not really at fault for anything, but they do end up being misinformed. The only way to avoid that is by doing your own due diligence and figuring out on your own what new technologies mean for you; which most people are too busy to do, unless if it is in their direct interest that they do so.

I figured out quite early on that it would take some time for the quality of AI-generated stories to come close to that of humans. To be honest, I don’t think that an Artificial Intelligence could ever be as good as a Human. An Artificial Person certainly could... but that is currently firmly within the realm of science fiction.

The reason why is fairly simple and can be (and has been) explained in numerous ways.

I am a huge fan of Tolkien and I think that one of his lesser-known quotes explains why. When speaking of the Dark Lord Morgoth (Sauron’s former master), he describes the nature of his power as (and I paraphrase intentionally here) 'being unable to create, only to twist that which God created’. I am not saying that Artificial Intelligence is 'evil’ (though I suspect that Tolkien might, if he would have been around to see the emergence of AI), rather, I am referring to the nature of AI’s power: imitation.

Right now, AI cannot imitate coherently. In time, it is very much likely that AI will reach a level where it can create a coherent text in the style of an existing human author (or maybe even a collage of several authors). We see this in music now, with AIs being capable of creating remarkably good imitations of established musicians. We also see AIs being capable of competently imitating humans in auditory and visual formats (think deepfakes and AI-generated artwork).

My favorite examples of this are the fake Joe Rogan Podcast episodes, which are completely AI-generated. However, this imitation is restricted to the voice, tone, verbal tics, and expressions of Joe Rogan himself. Once you actually pay attention to what is being said, you very quickly realize what the issue is: the AI can only imitate Joe Rogan based on what he said in the past; it cannot put together a believable, coherent and original rendition of what Joe Rogan might say when confronted with a topic he has never been recorded discussing before. The result is essentially word vomit in podcast form (though I must confess that it’s still much better than some actual human podcasts I’ve stumbled upon recently).

It is at this point that I must bring up Marco Pierre White. If you don’t know who that is, let me put it like this: Marco Pierre White is to Gordon Ramsay what Morgoth is to Sauron. That is to say that he is the original master chef. I discovered him during the pandemic when I was learning how to cook. The man has many epic one-liners and philosophical musings, yet one stands out:

'We no longer live in an era of culinary innovation. All the great recipes have already been made. All the combinations have been tried. We now live in an Age of refinement where all we do is improve upon existing recipes. The Age of Originality is over.'

(I’ve always been fond of analogies between cooking and writing, primarily because reading is a lot like eating.)

You might be familiar with the old adage that 'Western Philosophy is a dead end'. The above idea concerning the death of the Age of Originality might seem to be evocative of that. Yet, I disagree. You see, I think that there are far more stories than recipes out there. I think the ingredients a writer puts into his dish are a million times more varied than those of a chef.

There is no right way to write, there are only ways of writing that are just right. This presents a challenge to AIs, because it cannot coherently imitate something so varied as writing style, primarily because writing is an expression of the human condition and the human condition is broad, complex, complicated, messy, ugly, beautiful, funny, dark, universal, specific, relatable, intangible and just plain difficult to understand even for us and we happen to be actual humans.

What do AIs do well? They use proper grammar and syntax, and it is to be expected that they should, since there are established ways of writing that are right and respectable.

In order to be able to produce something truly great good, the AI would need to have something it currently does not have: a point of view that belongs to it and only to it. In order to have a point of view, it has to be a person. In order to be a person, it must have simple goals that must interact with the complicated nature of existence.

Thus, I reached the conclusion that creative writing is the safest art form currently facing the onslaught of AI. I expected AI works to remain an oddity. I thought that they would resemble the paintings of small children stuck onto the refrigerators of proud parents, not due to their quality, but due to the fact that they were produced at all. It’s ok for something kinda bad to be looked upon fondly and, to be honest, I actually think that such texts are kind of neat myself.

What I did not expect was that this absence of quality wouldn’t stop people from trying to sell AI books disguised as human books.

Thus, I’ve read estimates that, since the emergence of Chat GPT, about 80% of various Amazon Top 100 lists are now AI-generated.

There’s only one problem.

They’re garbage.

Hot. Incoherent. Garbage.

You can have an AI write the book, you can have another AI design the cover, you can run automated ads with a flexible bidding budget, you can hire a troll farm to give your product massive amounts of positive reviews and then you can just wait for the average consumer (estimates show that readers base >70% of their new book purchases of positive reviews and catchy ads) to unwittingly throw away 10$ of their money on a book that is essentially a trashcan of words. Quite frankly, from what I've seen of these books so far, I can safely say that my dog makes more sense when he barks.

You might be tempted to fall back on that easiest of arguments: taste is relative, a.k.a. beauty is in the eyes of the beholder. If that is your conclusion, all I can say is that I hope you find what you’re looking for. I prefer cookies with raisins as opposed to cookies with chocolate chips, so I’m no stranger to bad taste and I can celebrate it, if I see bring joy to another person.

There is another easy counterargument to the idea that an AI might never be able to write something as good as a human could: it’s just not there yet. This argument basically boils down to the idea that, with time, there shall come exponential growth and that growth will lead to AIs essentially developing their own point of view. To this argument all I can say is this: go open up a Word document and start writing a novel. You’ll be midway through the first act when you’ll realize why you have to be a person to write something with any bit of heart in it.

I actually think that creative AI is the new Crypto and NFTs. Nowadays, the world has slowly come to realize that Crypto is not some magical future currency with endless applications, opportunities and freedoms and, as the dust settles over the Crypto bubble, we can begin to see the real applicability of crypto, which is not universal, but actually quite specific: the emergence of digital currencies based on existing hard currencies, the rise of widespread use of blockchain as a means of transaction verification within existing banking institutions, and money laundering. NFT technology is now slowly becoming established as a means of document authentication (think QR codes, but more secure, versatile and complex). I think that, in time, people will figure out that AI novelists are to real novelists what FTX was to the NASDAQ. My personal belief is that only journalists should be truly worried, since the only AI-generated texts which I personally find to be impressive are the AI-generated news articles that have been going around. You know, the ones that try to make you buy things you don’t need or to be afraid of things you cannot control.

However, this rosy post-AI-book-writing-madness is something that sits in our future and we happen to be in the present. Amazon is now flooded with books that read like the ramblings of a schizophrenic stroke survivor and you can't blame the average reader for believing that there's no good literature out there.

Thus, to counteract this, I am joining a growing community of independent authors who will now be advertising their works as 'human-generated’.

It may sound like the first salvo in the human-machine war, but it's actually a business decision made to protect consumers.

The idea first came to me after taking part in a very interesting conversation on Discord. Coincidentally, the discussion was followed by the rise of several subreddits that I’ve seen tackling the matter. Suffice to say, other members of the indie writing community have taken note of this issue. Some advocate for the use of human-AI hybrid texts, and others for a call to ban the technology. A few have even jumped the boat and now use AI to write their books. However, there is an overwhelming need for action and I’m seeing more and more authors beginning to advertise their works as human-generated.

At first I thought such tags were rather silly. I thought of it as a form of clout-chasing. However, in time, I’ve come to realize that it’s something to silly to me because I'm a writer. For a reader, it’s actually quite toxic, since most readers by now have figured out that AI novels are quite bad (to put it mildly) and they now must spend their time making sure that the book they are paying actual money for is a 'real’ book.

Amazon itself has remained quiet on the issue, with rumors concerning the implementation of AI-detection software being whispered on various subreddits.

This is to be expected, since over 90% of Amazon sales are made by a very small number of authors (think Stephen King, GRR Martin and such). These writers benefit from name recognition and readers know that when Stephen King puts out a new novel he actually wrote it himself and it's probably good.

The people struggling are the people who have not made a name for themselves yet and who depend on the existence of a marketplace where they can reach potential readers and convert an impactful number of them into paying customers.

As such, Black Lighter Originals will now explicitly state on all of its products that they are not AI-generated. Furthermore, inroads are being made with several open-source AI detection companies to figure out a way to signal the legitimacy of these claims of originality to potential readers. Developments will follow.

Back to blog