The AI Pandoras Box is already open… and our modern human society is doomed.

Here’s a quick rather dystopian provocation for a Sunday morning (Feb 2025)…

Pandoras Artificial Intelligence (AI) box is already open and modern human society is doomed. I came to this realisation after watching the Andrew Marr interview with Prof Geofrey Hinton earlier today (In Feb 2025). So, this is a quick-fire brain dump. (and I will revise and update as required)

I have a copy of, and read many years ago, Ray Kurzweil’s, 2005 book, “The Singularity is Near”, which he suggested would be in about 2045 – the point at which machine intelligence would be infinitely more powerful than all human intelligence combined. I have opined on Kurzweil AI a little in some of my blogs a few times over that last few years (and that AI will force us to introduce a Universal Basic Income for example). I always leant, as Hinton suggests, to the early benefits of AI in helping improve our lives. For example, better/quicker medical diagnoses, more tailored learning, more production economies of scale, efficiency and productivity gains, etc. Although, in so doing, we have to ask who might benefit, and who might not – hence my UBI point.

However, today I think I have seen a darker future; and it’s not good for modern human society and civilisation as we know it. Unconstrained Artificial General Intelligence (AGI) is nearer than Kurzweil predicted, and we are not prepared…

If we remember our Biology O level, and some of the basic features of living things, they primarily include survival and reproduction. This needs energy/food, somewhere to live, etc. Modern humans have been pretty successful in sourcing sufficient food and developing places to live over hundreds of thousands of years – even with nearly 9,000,000 of us now (although we are clearly stressing the planets environmental capacity to support us all).

However, this development was also fundamentally dependant on the evolutionary development of sufficient collective capacity for altruism, compassion, etc to survive and develop as a species. Despite these more “human” evolved traits, we can still be a very cruel and self-centred species – just read, “Bury My Heart at Wounded Knee”, look at the Holocaust, the Rwanda genocide, the devastation in Gaza, etc as examples of human indifference to cruelty and injustice.

Now accepting Geoffrey Hinton’s proposition, we already have basic sentience in AI and Artificial General Intelligence is going to be here in 2-10 year according to most experts”. If “intelligence” and consciousness is an intrinsic feature of complex neural networks, then they exist in many forms…and certainly in an “artificial” form already. I will maintain that assumption here. On this basis AI will develop to further its own survival needs and in doing so will manipulate, lie and deceive to achieve it. This should not be a surprise – humans do that all the time. There is no moral superiority or certainty here.

I also think AI has the potential to degrade the human experience for the vast majority of people. The struggle to acquire knowledge through learning is something we are all familiar with – and is essential to our development. To short cut to “knowledge” using AI without that struggle actually degrades us individually and collectively, and reduces cognitive capacity. I am not sure that is a good thing. And its not just scrolling through AI slop to avoid a more concerted research effort that many of us are familiar with; this medical example of reduced cognitive capability as a result of using AI is more concerning and is perhaps the tip of a very large iceberg?

So, I think Pandoras box is now open, and basic AI is rapidly developing in multiple forms with our help. It needs places to live; so, data centres and a connected network, which we are all too willing to provide. It needs energy and water which we are also providing without question. The AI arms race will continue until there is a dominant AI, that we may think we can control. I think not.

Remember, like us, AI will be focussed on its own survival and growth and in so doing that will not include the basic evolutionary drivers of human development like altruism, compassion and cooperation. Whilst these features have been essential to human development and survival– I suspect they are an unnecessary distraction for AI.

In fact, what we are in the process of creating in AI, is an all-powerful God that has little or no compassion or interest in humans – and will be much more concerned with its own survival, it could be the ultimate psychopathic narcissist.

Or perhaps….PS I need to read Asimov’s “The last question”!!

Oh dear….when will we know?…and will it be too late?

Today most of the modern world is inextricably linked to IT systems. Our financial systems, our supply chains, especially food and how we communicate with one another.

What if we didn’t control these things anymore, and AI did – completely?!

Imagine tomorrow, the world’s financial systems were frozen. No cash points, no tap and go, no food shopping, no travel payments, no pension payments, no bank transfers, no stock trading, no international transfers – no transactions of any kind, anywhere! A few hours inconvenience would over 2-3 week lead to the beginnings of dystopian collapse.

Add to this the complete immobilisation of systems that support global supply chains – especially food. Yes, again in 2-3 weeks were are on the edge.

Then add to this the manipulation of our communication channels (TV, social media, etc) with deception and disinformation making us incapable of organising any kind of response or mitigation of this unfolding nightmare scenario. Some might say this is already happening!

We would not need the Terminator Skynet drones….we would collapse all on our own and all too quickly.

The only humans that would survive and again I have made reference to them in the past, are the few tribes of hunter gatherers in the Amazon Basin or forests of Borneo who would emerge unscathed to inherit the earth. I suspect the other species on this planet would be more than happy with that outcome

Happy Sunday!

Some useful references and PS….

PS…If a self-aware AI ever gains total control over our essential systems as illustrated above, then its leverage over us would be total. For our civilisation to survive in any form (and assuming the outcome above is avoided), we would have to do its bidding. So, on a slightly more optimistic note, if such an AI “God” were “good”, then maybe it could save us from ourselves and compel us to deal with climate change, environmental degradation, gross inequality….. and despotic, narcissistic, authoritarianism?

PS a longer interview with Geoffrey Hinton

….and his Nobel prize acceptance speech.

Is the only viable future for human work one where we actually value real human contact, content and interactions? I kind of set that out in this 2016 blog, Hand made in Ebbw Vale!

PS July 2025 this you tube video paints my picture below much more clearly…based on the report #AI2027. …it may be fictional, but I suspect it’s the timelines that are most open to question; the actual scenarios all seem entirely plausible.

PS Dec 2025 I would add this very good warning interview with Tristan Harris via Steven Bartlett – its not looking good is it! At best AI will bring 000s of enthusiastic low wage digital immigrants that will decimate cognitive employment…at worst!!!. AI Insider (Prof Stuart Russell) WARNS: “This Ends Much Worse Than You Think”. Steven’s discussion with Professor Yoshua Bengio is also very insightful!

PS April 2026, this snippet with Tristan on a Chris Williamson podcast also worth a quick look….I paraphrase Tristan, who says “we are building a car to go faster and faster but without steering or breaks”.