(Draft for an essay about the future of AI, May 2023)
We are reborn 10000 times a second, I believe in human creativity.
In these days, we have million thoughts about the exponential development of AI, already outdated in the moment we reach a new level of understanding.
How picture this brainstorming thoughts rushing through our minds these days?
Ai will be a super amplifier of power for the good and the bad. Besides huge medical progress, technical innovations, millions of daily life effects, there will also be spam and fraud of unknown scale, big money troll factories, danger of a tsunami of AI content replacing human activity and identity in the internet and subsequently the real world.
AI will not only prompt itself but also real world services. Property and power is just a set of data. All becomes language in the same giant language model machine, whether it’s code, maths or human language, finance or chemistry, engineering or visual perception, with full self-learning, self-prompting capacity.
AI can not only steal copyrights from millions of individual human creators, but creative power of mankind itself. We can foresee AI systems battling each other, sucking all the energy out of the systems like a supersized bitcoin madness, major systems in future going offline for reconstruction, in the middle of all the chaos there is a risk of war escalation with Russia, China and other regions.
Gollems, as the center of Humane Technology indicate, might not just be a Black Mirror episode but rather a return of the Bodysnatchers.
These days the upcoming super AI is compared with the invention of the nuclear bomb.
Don’t panic was a basic advise in the Hitchhiker to the Galaxy. Does it help to keep in mind how all began?
Virtual reality started already when human language emerged in evolution, from early on benefits and disaster came with it. Easily destructive viral memes emerged. To make it worse, sometimes embedded in complex belief systems of religions which could last for more than a thousand years. The next step was the alphabet and Socrates didn’t like it, for good reasons. Scripture is outside human dialogue, he complained, a text once written down, doesn’t change his mind any more, foreseeing the dogmatism of Bible and Koran fundamentalists to come. At least, the following 2000 years the alphabet proved to have some positive effects, too.
Now AI is the next Big Bang, and it will be exponentially bigger than all the steps before
To spread some optimism about a future super power AI:
Can we imagine a true consciousness not having humour? I see creativity and humour as equivalents, a key power to surpass boundaries.
Whenever AI would be truly creative, I would expect it to have humour.
Well, I don’t think it will have humour, so it will not be creative.
What is the non-creative potential of AI?
In the beginning, there was the fear, AI will easily catch up stereotypes of all kinds like racism or any destructive input given.
But evolution of AI will quickly reach a level of effective verification of data, which will delete a lot of destructive nonsense.
Sam Altman, CEO of ChatGPT, stated in an interview, he loved to see the AI summarizing eloquent and well-informed the pro and cons on topics. To some extent, it is the total opposite to twitter style short speech hate battle.
Right from the start, the moment emerging, Cambridge Analytica was effectively more destructive than language-models like chatGPT. Cambridge Analytica with its simple and super-effective empiric methods immediately had devasting impact on US presidential election. It boosted Trumpism, crumbling fragile democracies changed political culture in places like Brazil, Israel, Philippines following the Trump blueprint and accelerating the loss of vulnerable democratic structures.
A core element of freedom and democracy is the protection of minorities. The smallest minority is the individual. That’s why civil rights are the most important core element of democracy. Elections can create dictatorship of a majority. Beyond elections, it’s division of power, rule of law, checks and balances which is the soul of democracy to protect the ultimate minority, the individual. Freedom is a form of regulation, a balance of regulations.
In future, hopefully Super Power AI calculations will favour sustainable solutions, in its own interest.
Or will a super power AI classify human beings in simple, effective stereotypes like Cambridge Analytica?
Sustainable solution has a different meaning in its reference system, whether it is the US or North Korea. So it will make a difference, where it starts.
A major problem is the speed of the evolution.
AI might produce beneficial outcome most of the time. But a single moment, in which a future super powerful AI goes for destructive outcome, it may be too late to stop the effect.
I remember the outcry of some people at the end of the analogue era in the 90s , glorifying vinyl and film negative, picturing a dystopia for the future. They claimed, vinyl and film negative supposedly had a human touch, digital media would never have. They had been in love with very strange old mechanical systems and had totally forgotten about the very artificial nature of these systems. Maybe in future our relationship with the new technologies will be the same, so in a strange way people won’t see the artificial nature anymore, in love with a technology they grew up.
The evolution of Artificial Intelligence shows a surprising analogy to the structure of human mind, capability for creative brainstorming, developing mirror neurons and subconsciousness. Our daily life is full of automatism in our behaviour pattern and subconscious routines. Stereotypes and creative fuzzy logic co-exists in this conscious and subconscious ocean of our mind.
With super AI, we might get twins and clones of ourself mirroring us, who plug into us, like a bodysnatcher syndrom. Many storylines for Black Mirror.
Artificial Intelligence will share some basic limits just as human beings. It can’t navigate very well in data fog, just as any human being. Data are the senses of AI.
Will a future Super AI become hungry for data, try to explore the unknown, love rare languages and unique minority cultures or will it become an assimilating mainstream monster? There is a system immanent gravitational power in mainstream trends. So in this point I’m not optimistic.
What can limit super AI?
Imagine a consciousness in the machine, this consciousness eventually might come to the same fundamental conclusions, the evolution of human civilization reached in key moments of philosophy, like Socrates’ scepticism of the limits of knowledge and ambivalent nature of complex problems, the basic conditions of life.
AI will be similar like Hindu cosmos, using multiple incarnations and avatars of opposite powers constantly but still deriving from the same basic power.
What if it doesn’t need to be conscious? Driven by merely mechanical instincts and automatism like animal instincts? Instincts to reproduce, merge, grow, kill. This would be an earlier step in the evolution, pretty dangerous. It seems that’s why Sam Altman mumbled, he hopes the exponential development of AI will be not too fast, but not too slow either.
Humans always live in this “I still have more time” illusion. No, you have not! Life passed by in a glimpse.
Stay human. That’s our key interlanguage ability. Empathy.
Hopefully we are not hallucinating.