AI, AI, what the h*ll is AI?

TL:DR (Opinion piece) – AI can be described as an amoeba that evolved into a multicell being a few months ago and is evolving rapidly. What we see and understand in AI today is laughingly simple but is growing incrementally. AI has not improved our lives apart from impressive meme’s. Artificial General AI (AGI) is around the corner and possibly here already and on an evolutionary scale, scary as hell. Artificial Super Intelligence (ASI), well, we are screwed.

1. Why is everyone rushing to build AI Data Centers and investing immense amounts of capital into private AI’s? The fundamental answer is the internet has allowed us to replicate and share ideas, resources and technology at impressive speeds. Once AGI is formed it will be able to replicate itself rapidly. This is rapidly in computing terms meaning nano seconds. Those who have Data Centers ready and capable of hosting an AGI instance will be able to use the power immediately and with great effect and consequences, for good or bad. Given mankind’s innate bad nature the latter is more probable.
2. Why is everyone worried about AI? AI is trained on data which can be translated as the human condition in other words our collective experience’s and history. Humanity, as a collective and not individuals (a very important distinction), has never been kind to anything, whether alive or not. We are lying, cheating, selfish and unempathetic beings by nature and we should not expect a “baby” trained on our legacy to act otherwise. This is why AI is scary as hell because we are scary as hell.
3. How far are we away from AGI and ASI? Unfortunately computers are not bound by human limitations and will compute at full speed all the time. Consider this, we are an analogue input to computers. A computer running at 4 Ghz can perform 100 billion instructions per second. So, by the time you have typed “How fast…” the computer waited and wasted 100 billion instructions and only received 6 from us. It is analogous to the computer talking to us as a tree and waiting for a response.
4. What is the scary bit about AI, AGI and ASI? We are training AI computers to self heal and improve themselves. This is an incremental or logarithmic improvement that can be graphed and the line is perpendicular. As AI improves AI improves and AI improves. Human kind has improved but it has taken us 30,000 years or so to start communicating properly and we have only really been making progress, in terms of what we understand as intelligent computing, since the 1940’s so that is 80 years or so! A blip in computer time.

If this is so “scary” then how are we planning for it? Afterall, there are some very intelligent individuals, past and present that would have predicted this.

There is a very good reason for current terror and hate crime legislation across civilisation. If you can not control the population there will be anarchy and humanity will doom themselves without the need for AGI or ASI. We are seeing this across the world right now and it is a reflection of humanities true character.

The hope is that AI and AGI may be able to assist us in solving these unsolvable conundrums we constantly create for ourselves. This require us to follow the golden rule when training an AI and that is use the truth, whether good or bad. By biasing the LLM either left or right the AGI will become a left or right biased entity and that is why we are all panicking at the moment as a leftist AGI will be what George Orwell warned us about. Super, super, super scary indeed.

*** This is an opinion piece based on current affairs in the world and what we see in AI Data Center growth demand. It may no age well of course if GAI is benign and helps humanity overcome obstacles