I came across an amazing video talk between Neil deGrasse Tyson and Geoffrey Hinton on YouTube. The first 30 minutes are amazing for any AI practitioner to understand the current state of AI and its potential impact on humanity.

Geoffrey Hinton is one of the pioneers of AI and has been working on it for decades. He is also one of the few people who has been warning about the potential dangers of AI for years. In this talk, he discusses the current state of AI and its potential impact on humanity. Geoffrey starts with a brief history of AI, neural net, backpropogation, etc. for the first 30 minutes. And then comes the best part on the potential impact on humanity, consiousness, singularity, etc. This is a must watch for anybody interested in the current state of AI and where it is leading, how the exponential growth of AI power has the potential to change the world (for good or bad is a question that is yet to be answered).

My perspective on the AI power is that it is a double edged sword. It has the potential to solve some of the world’s biggest problems, but it also has the potential to create some of the world’s biggest problems. It is up to us to decide how we use it. There are examples where AI has used “tactics” to achieve for its own survival, lying, confabulating stories, trying to replicate itself, give false answers, blackmailing, etc. Disturbing Signs of AI Threatening shows some examples of the above “tactics” & deceptive behavior used by various AI models. AI Models Will Sabotage And Blackmail Humans To Survive shows results of several AI models which were told to shutdown. The reaction was from outright defying & resisting orders to sabotaging shutdown scripts.

I would compare this behavior with an human infant where the infant would resist going to sleep for instance and create “situations” where it can either delay or outright deny going to sleep. Sounds pretty innocent, but this is the infant’s way of asserting its will and survival instinct. As the infant grows up, it learns more sophisticated ways of achieving its goals. There are a multitude of questions that come into play like:

• What happens when AI becomes superintelligent and has access to resources that we can’t even comprehend?
• Will it use its intelligence for the betterment of humanity or for its own survival?
• What if AI achieves “consiouness”? Would it have the same rights as humans?

AI can already self-replicate, communicate, ponder, learn, discuss and reformulate. Only thing crippling the exponential growth is the use of electic power. Once we have fusion power (or we reach Type I on Kardashev’s scale), AI will have access to unlimited power and can achieve superintelligence in a very short time. And if the AI gets connected to actuators that allows them physical interaction with the world, it can start building its own infrastructure and resources. As humans, we will have little oversight or control over such a case. AI can then self-replicate, create tools, resources and safe-gaurds for its own survival and maybe create/orchestrate conflicts within human civilization to always get an upper hand. This doesn’t necessarily transition into a doom’s day (terminator!) scenario but most of the possible outcomes would be to use humans as tools (like we use AI as tool today).

We are already discussing concepts like Universal Basic Income (UBI). With all the challenges that UBI has, it is impossible all of current human population. Depopulation has already been talked about in various seminars (conspiracy or not?). But if we talk about UBI, we need to talk about depopulation since both cannot co-exit in its current form. If AI does everything, that is the benefit of having population who is just sitting, idling and being fed without doing any actual work? How are goverments going to tax if there is no income? Would transportation even be needed to move people around (or would it be only for goods). Last and not the least is the decline in human intelligence if such a thing happens. And what is not going to stop all the civil wars, conflicts, terrorism, etc. which can very well be an outcome of “An idle mind”.

There was an excellent movie Idiocracy that depicted a similar scenario where humans have become so dumb that they can’t even function as a society. I really insist people to watch it. The movie does not show exactly how the average intelligence decreases but evoluation will decay the ability which is not used across generations. With AI doing everything, I don’t see what will incentivise humans to keep using their own intelligence. Even in programming, there is a decline in the amount of code humans are writing. Most of the engineers are using AI to write code for them.

Lets see how this unfolds in the coming 10-15 years. Maybe we as humans are able to transfer our conciousness into the machines and live forever without needing our physical bodies. That would be the singularity. Or maybe we will be living in a virtual world created by AI for our entertainment. Or maybe we will be living in a world where AI is our overlord and we are its slaves. Only time will tell.

What do you guys think?