The Amazon Echo image credit – digitaltrends.com
This is the last part of our special series on Artificial Intelligence (AI). If you are catching it for the first time you can check out our early posts here, here and here where we build on the bigger idea behind AI.
In this series, I have seized moments to highlight just how quickly AI is unfolding. The interesting bit is that the developments silently running in the background, like computer processes, for the most part, are away from everyday headlines but it’s still a big deal.
From my point of view, judging by the number of AI startups that have come up and are being acquired, we are in the middle of an AI gold rush right now which some have characterized as the AI spring .
This hurry is also the reason behind key backers and AI enthusiasts such as Elon Musk, the CEO of Tesla, have been a sounding alarm.
Roboethics and you too will soon make your own robot
The head rush to explore the limits of AI is arguably leaving a huge door open to all kinds of things that some genuinely worry will lead to the end of humanity as we have always known it.
In response, there have been ongoing calls to arrive at some sought of AI code of conduct, roboethics, or ethics for artificial intelligence. The purpose being to stir developers away from intentionally harmful AI systems.
The problem with these codes of conduct is that AI is increasingly becoming accessible to just about anyone who can code. In many ways it’s like attempting to police the internet. Most open source AI modules and services are sitting and will run from the cloud.
I was recently at the the Zimbabwe Centre for High-Performance Computing (ZCHPC) located at the University of Zimbabwe.
I found university computer science teachers from all over the country in training on AI systems development, with the primary goal of being to triggering interest that would cascade down to the respective local universities.
What this means is that barriers to entry are rapidly falling, it will soon become commonplace for just about anyone with the passion and spirit to code and create all kinds of AI solutions.
Since the 80s and 90s we have been buying components to build PCs at home. In the coming years, barring regulation, it will be possible to assemble components and build a self-aware reasoning helper to weed your garden or walk your child to school.
In the last article, I talked about the current but certainly being overcome problems with missing fast enough computers, too big and too expensive to facilitate the arrival of artificial general intelligence (AGI).
AGI represents a stage in the AI timeline were machines gain general-purpose intelligence.
This is the stage were the same robot, for example, can cook your meal, have an emotional chat with you and drive a bus full of tourists in mountainous passes.
The current estimate from AI’s leading lights is that by the year 2030 the world will welcome computers at a general scale with the same computing power as the human brain.
Road to ASI – we are becoming ants
If machine intelligence becomes common place by 2030, where you have robots and systems possessing human-level intelligence, the next step on the AI timeline is combining all that machine power into something greater and more powerful than the human brain.
This has been referred to as Artificial Superintelligence (ASI).
ASI will be easier to achieve because AGI machines will double themselves in rapid succession getting ever more intelligent each time they create copies of themselves in a similar fashion you cluster PCs to make them powerful.
What it essentially means is that from that point on, human beings will essentially become subservient to super-intelligent machines. How much more intelligent these machines will be has been sometimes put at a million to a billion times more intelligent than humans.
To get a sense of it. Think of an ant in your backyard. Do you care about it or even stop to think it exists when you wake up in the morning?
You certainly don’t hate it but you just don’t care about it. You are not bothered what happens to it if it rained or thundered. ASI, it is feared, will have the same attitude towards humans.
This is the reason those who oppose or at least worry about the future of AI, call it our last invention.
When and should super intelligent computers arrive, in their eyes, humans will become just like the ant in your backyard in comparison to a billion times more intelligent self-aware machines.
Further worry is that there will be no button to switch AGI off let alone ASI, just like no ant knows how to find your main switch in your house and dis-empower you.
Science or fiction? – My own conflict
I admit I am somewhat conflicted. At this point, a lot of this stuff sounds fictitious. But what is certain is that this is real science.
It only takes a deeper look at what is actually happening in AI research and development that you will soon realise tech is on a path to something never imagined in power and capability.
I do predict, though, that at some point soon, all hell will break loose as religious leaders and consequently regulators will begin wading into the debate with the aim of taking bigger control of the future of AI.
In this series, we have traced and presented the idea driving artificial intelligence, from basic AI to AGI and then ASI.
Having set out this important context, my new focus will be exploring opportunities in our Zimbabweancontext, at least as it pertains the benign aspects of AI.