AI – Imitation is the Most Insincere Form of Flattery
This summer, I’ve been thinking about AI – Artificial Intelligence. This is AI’s Year Zero. Hence, this article attempts a response to AI by taking an informed view of the current issues and how these may impact writers like me.
I use digital technology daily as a consumer and content creator. After all, the origins of this website date back 25 years. And who doesn’t use the Internet? However, I no longer use social media platforms. This is partly because of the declining quality of content and because they extract behavioural value from us for profit. As usual, it’s not the technology itself that’s harmful; it’s how it’s used (and abused).
In short – I’ve decided to use something other than AI on this website – me.
“Look, Dave, I can see you’re really upset about this. I honestly think you ought to sit down calmly, take a stress pill, and think things over. I know I’ve made some very poor decisions recently, but I can give you my complete assurance that my work will be back to normal. I’ve still got the greatest enthusiasm and confidence in the mission. And I want to help you.” 1.
Claim and Counter-claim
Proponents of AI, usually those with the biggest vested interests, claim that it’ll save the world – from climate emergencies, diseases, poverty, and ourselves. In short, Utopia is, like always, just around the corner.
Others predict dystopian outcomes, where AI enslaves us all to robots, with the inevitable extinction of the human race. Maybe – but AI is NOT sentient; while AI now does well in the Imitation Game, it remains an inanimate parrot and wouldn’t pass the Dunning-Kruger incompetence test. If AI were to develop a mind with consciousness and thought, that would be genuinely scary.
In any case, all these claims are media smokescreens, a diversion from the actual impacts of AI that are already happening. Grandiose claims are different from the intended goals that Big Tech wants from developing and owning AI. The real Big Tech goals are the same as ever: the corporate extraction of wealth and power.
The State We’re In
Almost everything humans have published (including the content of this website) has now been “crawled” by bots. That enables search engines to do their work and can also create vast machine-readable datasets. All these datasets are used to “train” AI machines, which must be repeated many times to “optimise” potential results. Additionally, even your data held in particular cloud services has probably been used, whether private or not.
Technology models then manipulate this data with massive computing power, predictive algorithms, and neural networks. With language models, the machine composes text or speech answers to questions asked of it. It does this by making statistical predictions about the most likely word to occur next in the sentence under construction. The more training that is undertaken, the more the likelihood of generating a plausible result. It’s not necessarily accurate or truthful, but it is often plausible.
Similarly, this applies to other creative disciplines such as art, photography, performance, music and films. In short, enormous amounts of data are chewed up, stored, and spat out by computers in Big Tech server farms. We’re talking zettabytes worth of data: total annual internet traffic reached one zettabyte in 2015; this year, it will be 103 zettabytes.2.
Real Threats Now
AI also needs continent-sized amounts of electricity. Hardly eco-friendly. A recent study suggests that AI uses 2.5% of the world’s electricity. Some of this is renewable, but carbon offsets make any green claims highly suspect. As for the water needed to keep data centres cool, that’s massive, too. And resource demand will grow as AI encroaches. This is an aspect on which Big Tech keeps silent.
Moreover, Big Tech needs to be more transparent about most other aspects of AI. While researching for this article, I discovered that some “training data” fed into the Big Tech machines has been undertaken in the third world with exploitative conditions and low wages. And this work even takes place in refugee camps! Some workers have developed PTSD from the material used.
Imitation is the Most Insincere Form of Flattery.
The insatiable maw of AI has no moral compass. It doesn’t know the difference between “good” and “bad.” It demands (and produces) appalling graphic sexual violence, hate speech, and abuse just as readily as it wilfully consumes the novels of Jane Austen and the plays of Shakespeare.
You might find it harmless fun to read a wine review that’s AI-written in the style of Stephen King. However, would you like to see what appears to be a family member victimised in a horrific abuse video? Or being scammed out of your life savings using what sounds like the voice of a trusted friend or relative? “Deep fake” is already here, and AI is making it even harder to spot. How, then, will we distinguish truth from fiction?
AI is only as “effective” as the data it feeds on and can’t distinguish between truth and lies. With restricted data, it produces obvious errors and garbage, which Big Tech calls “hallucinations” to give you the impression that AI is somehow alive. Gubble, gubble, gubbish, as Philip K. Dick would say. Another example: Amazon sells books written entirely with AI – those AI-written books about Wild Mushrooms are so error-strewn that Fungi experts consider them positively dangerous!
Datapanik in the Year Zero
Though only in infancy, AI offers unregulated, opaque, and unfettered mimicry. The data used for “training” such programs has often been obtained illegally, flagrantly breaching copyright and representing theft on a massive scale. For example, it was recently revealed that Stephen King’s novels had been taken by AI without his consent, acknowledgement, or payment, alongside 183,000 other books.3. In the wine world, three Jancis Robinson books are included in that dataset – how do you feel about that HRH?
As is common in Big Tech, the attitude is “just do it and ask for forgiveness later.”
Your Replacement – a Toaster
Millions of jobs across every sector risk obsolescence. Previous automation at least replaced aspects of tedious and repetitive work. Now, AI threatens to replace decision-makers and entire creative industries. These will become obsolete as it is cheaper to serve up simulacra without imagination rather than employ a human to make something new. Meanwhile, search engines (Bing, Google), Microsoft Office, Google Docs, and even email incorporate AI, as do translation programs and grammar/spellcheckers. Our participation in AI is becoming ubiquitous. However, this is often without our knowledge, let alone consent.
Creatives, whether in writing, acting, music, art, or photography, need to stand together. They are coming for you. And there’s more – AI imitation can easily pass written exams and write research papers – so in the future, what value are your qualifications? Or, indeed, your education?
Writing in General
Writing, whether exciting or dull, has, until recently, been an entirely human process. At best, it’s a mark of creativity and intelligence that distinguishes our species from the wider animal kingdom. So, what impact does AI have on authors and, by extension, their readers? There are authors, poets, and journalists at risk, sure. But it threatens an entire ecosystem of academics, illustrators, translators, and editors.
To quote The Society of Authors,4. “The UK government makes much of the £3.7 billion contribution that the AI industry made to the UK economy and the 50,000 people it employs. This is just 3% of the £115 billion value of the creative industries, which employ some 2.4 million people.” On that metric, Big Tech should not be calling the shots. However, another way of looking at it is that there’s plenty left for AI to gobble up.
Wine Writing
Wine writing is not exceptional; its limitations apply to the entire genre. Is it any significant cultural loss if generative AI and chatbots replace wine writers? Probably not. But that’s not the point. AI replaces my genuine, if time-consuming, human attempts at creativity and purpose. And with what? Machine text, whose advantage is that it’s quick, cheap, and even duller.5. Soon, wine writers, educators, and even sommeliers will cease to exist.
Meanwhile, “member organisations” (regardless of speciality) that have not alerted their members to AI are doing them a grave disservice. Like the Dodo, they will soon become extinct because all their members will be.
Hence, I’m grateful that the Society of Authors (SoA) is a member organisation taking a clear and active stand on AI. It’s undertaking research and lobbying governments and big tech companies to implement effective international controls. These would protect jobs and incomes and enforce already existing legal copyright, creative, and privacy rights. Meanwhile, the SoA offers practical contractual advice to help authors protect their commercial interests.
This website will continue without AI and despite AI.6. But only because I do it for love. Try simulating that.
And Finally
Inevitably, the AI genie is out of the bottle, and there will be winners (a few) and losers (most of us). Inevitably, the impacts of AI will affect everyone across many aspects of our lives, whether intended or not. There is the potential for good outcomes, but only if action is taken to control its use.
As it stands, AI is the latest example of unregulated mass extraction of value by Big Tech for their sole profit. It is not free; it’s plagiarism. You and your creativity are the product. We serve it, yet it should serve us.
At one time, Google used the phrase “Don’t be evil” as its motto, including it in its IPO share-offer documents and Corporate code of conduct. It doesn’t use it anymore, so make of that fact what you will. However, I recall a famous phrase: all it takes for evil to flourish is for good people to do nothing.
AI is plopware.
Notes
1. Spoken by HAL 9000, the rogue spacecraft computer in Stanley Kubrick’s 1968 film, 2001: A Space Odyssey.
2. A Zettabyte is 1021 bytes (1,000,000,000,000,000,000,000), or a billion Terabytes, or a trillion Gigabytes.
3. Check out this database on The Atlantic. Stephen King is just one of many published authors. However, as the “King of Horror,” it seems apt to mention him for this article.
4. Source: The Author, Summer 2023.
5. To paraphrase Frank Zappa, AI is for “people who can’t write, interviewing people who can’t talk for people who can’t read.”
6. The Wine Alchemy Policy Page stipulates no AI usage is allowed on this website. Specifically, GPTBot and CCBot have been barred from crawling the pages of this website because they scrape data for AI.
For more about the beneficial uses of technology, please look at my pieces on Blockchain.