The Imitation Game

Alan Turing was a trailblazer in many regards, there are a few technology related terms that he lends his name to, the most pertinent of our time is the Turing Test - specifically the Imitation Game whereby a human uses a computer to engage with two entities at once, one being an actual human and the other being a machine, with no indication which is which. The machine is said to pass the test if you as a human can't tell which is the human and which is the machine.

This is perhaps the most pertinent of our time because of the rise of AI, or to be more precise, the rise of the Large Language Model [LLM] which is a complex algorithm that uses machine learning to process a large dataset consisting of natural language and then mimic that language. The reason this is distinct from AI is because of the lack of reasoning, something which the newer variants of most of these models are now incorporating and developing at pace. The term "AI" in its purest definition has mostly been abandoned, held onto by Computer Scientists with the same pedantry as the distinction between "The Internet" and "The World Wide Web" which are not the same thing but for all intents and purposes now most people use the terms synonymously and any effort to hold onto that distinction seems pointless at this stage - AI has gone the same way, whether it is "technically" AI or not is now irrelevant.

What is relevant is that there is a now a deluge of content written by AI that is creeping into our lives. Some of it is "raw" being the unfiltered output of an LLM, whilst some of it is "revised" being edited by a human who uses the LLM output as their basis and builds upon it. The former is a lot easier to spot, the latter is a little more difficult to see because it does incorporate some human sentiment and nuance in the minor revisions made during the editing process.

If you've been following this blog for a while, or the last two months at least then you might be familiar with my 52 Goals for 2025. One of these goals was to read more history books and learn about countries whose history I have largely been ignorant to. My previous education on the subject was heavily Eurocentric being from the UK and also heavily skewed towards our own domestic history and the various interactions we've had with our neighbours. To that end I have been reading a lot of books and it's becoming painfully obvious that human-written content is already proving harder to find when it comes to anything factual and relatively easy to research - that's not to say the LLM generated content is necessarily accurate, that leaves a lot to be desired.

I still hold onto my view that I want the AI bubble to pop, so that real people can be left to find real uses for the technology in the absence of corporate interests, the same way the dot-com bubble popped leaving the internet largely to the mercy of normal people which allowed it to mature. I don't think that's going to happen though, not because of a lack of disillusion, there are certainly enough people who are already tired of it, but because of fragmentation. The early monopoly of OpenAI has been broken, everyone and their dog is sticking an LLM in their products and services and the emergence of DeepSeek, an open source LLM with reasoning capabilities mired in controversy has only accelerated that fragmentation - the controversy being, ironically, that DeepSeek allegedly trained its model on ChatGPT without permission, not unlike ChatGPT being trained on artists creative works without their permission.

The other reason I don't think that bubble will pop now is not just that it's no longer a single bubble, but that governments have begun saying the quiet part out loud, that they see AI as a weapon and openly admit that and want to be the ones to perfect it and use it first.

There's a quote from Doctor Who from the episode Day Of The Doctor when the War Doctor steals 'The Moment' a deadly weapon so advanced it developed a conscience, the General asks “How do you use a weapon of ultimate mass destruction when it can stand in judgement on you?” that is what we approach, if a super intelligence does emerge and it is weaponised, those who attempt to wield it will no doubt get more than they bargained for, but any hope that it might save us all from oppression is slim. Any judgement of humanity is likely to be collective, just as disease ignores borders and nations, human concepts of division that separate "us" and "them" won't hold any relevance to a super intelligence if it did emerge.

There is a concept called Moral Relativism which holds that morality sometimes permits us to stomach certain acts depending on who carries them out, that our enemies can't behave in a certain way e.g. murdering our people, but we can, murdering their people, because the motivation is different, the act itself is not the determinant of morality, one is offensive and one is defensive and moral relativism allows for the latter.

There is another concept called Moral Particularism which is where humanity's cruelty and the depths of the horrors of our past can be found. Moral Particularism doesn't define actions or even motivations as factors in whether something is moral, instead it defines morality in terms of who performs the action, "I can do this thing but you can't", "It's right when I do it, and wrong when you do" - the motivations can be exactly the same.

The best examples of this moral particularism are found in imperialism, throughout history from its most nascent forms in terms of tribalism, to the structured imperialism of the not to distant past, there is a desire that persists to rule one and all not for the sake of unity but for the sake of power and control. It rears its ugly head time and again and even when countries ostensibly abandon imperialist desires and rhetoric, dig a little deeper and the reasoning is usually the recognition of some greater power whose ire would be stoked if your ambitions grew in that area - in other words the imperialist tendencies of nations are kept at bay by a stalemate, not a lack of desire or an abandonment of that base nature.

The root of all this conflict stems from resource management. Whilst its tempting to label money or religion or some other motivator as the factor, those are all symptoms, systems designed to enable power and control and more importantly access and consumption. Resources are the underlying cause, and the science of their management and the decision making that surrounds that management - Economics - whilst often bundled with Finance, the two terms are distinct, and economics existed for as long as there have been resources gathered, long before we had a word to describe the behaviours of hoarding them and consuming them. In the timeline of human achievement, Finance is a relatively recent invention.

The ultimate Turing Test, the absolute Imitation Game that will eventually play out should a super intelligence emerge, will be how akin to human behaviour that intelligence behaves. If it judges humans the way we judge each other it will recognise that humanity would never accept being "beta" you only have to look at the wealth of science fiction written by humans before it came along to see how much fear and anxiety the human race has instilled in the idea of something becoming more powerful than us. Look at history and how we have behaved when confronted with our subdivisions, our borders, our nations, our states, and the response they had to others becoming more powerful and the writing is on the wall.

We're all so worried about telling the difference between an AI and a Human when the real worry is whether an AI will act like a Human and all that entails given half the chance.

Live for today, there might not be a tomorrow.