Future of AI

Best way to predict the future is to create it – that’s a cute saying but in our reality soon the future will be created by AI not by humans. AI revolution is upon us and this saying no longer is valid. Furthermore not only we don’t know what future AI will do, but we don’t even know how contemporary AI works: case in point: AlphaGo Zero.

AlphaGo Zero is a proof that future of AI is unpredictable: AlphaGo Zero is Google’s Artificial Intelligence to create AI that beats humans in playing ancient Asian game of Go. Until recently all researchers were predicting: game of Go is much more complex than chess, so it will be another 40 years until AI can beat humans. But here comes something unpredictable: Google’s AlphaGo beaten all human players already. Interestingly it self-learned by playing with itself – by using so called “reinforcement learning”. Furthermore human Go players say that AlphaGo plays unlike humans do, that it is not “thinking” like human Go players do. This is foreshadow what is coming in future AI: AI that we humans neither can understand nor we can predict.

AI winters are over. AI winter is a term used by AI researchers to denotes periods of time when investments in AI dried up and there was not much developments and news. If we live now in AI spring, what would you call future of AI? AI summer?

Humans think that animation of machine – how much they feel to be human – is directly correlated to their intelligence. But that’s incorrect. Tech-animism – humans attributing unconsciously (not subconsciously as Leonard’s mother pointed out in latest Big Bang Theory episode) humane features to voice assistants like Alexa, Siri, Google Assistant, are very misleading: for example more natural tone of voice assistant makes it more intelligent to humans even if given voice assistant is less intelligent.

In near future AI that we will meet the most will not be humanoid robots but rather voice assistants embedded everywhere and intelligent features of devices and services.

Self-driving cars are unavoidable. Probably it will take till 2030 until self-driving will be wide-spread and de facto standard though: there are many problems to be solved yet and wholly new digital infrastructure will be needed for these cars to spread.

Speaking of infrastructure for self-driving cars: do you know what V2X is? V2X is Vehicle-to-everything communication: passing of information between vehicle and everything that might affect the vehicle. Funnily enough this new communication systerm encompasses also: V2I (Vehicle-to-Infrastructure), V2V (Vehicle-to-vehicle), V2P (Vehicle-to-Pedestrian), V2D (Vehicle-to-device) and V2G (Vehicle-to-grid). So you can see that self-driving cars will not be merely about cars and sensors embedded in these cars, but also about whole environment, whole system and infrastructure. You can easily imagine that in future self-driving cars will be communicating for example with smartphones of pedestrians to ensure that car accidents will be minimized. And if not with smartphones then with GPS badges placed on people and animals. And from GPS badges there is only small step towards implants that will be embedded in human body and that will enable bi-directional communication between humans and AI in computer space.

While in the past visions of future where “big brother is watching” you were widespread, now it’s not a future anymore and big brother is AI or algorithms of big corporations. AI revolution will bring massive changes in private sphere of people, in privacy. Noways 3-letter agencies from USA can practically use any web cam in any laptop to observe you, so you should cover these cameras with stickers. For fear of constant listening some tech reporters refuse to own smart speakers like Amazon Echo, Google Home or Apple HomePod, but one episode voice assistants will be embedded in many devices, this will not be a choice. In era of AI revolution privacy is a goner.

Day job? You need a job? Well, BCG (Boston Consulting Group) predicts that mid-term (about till 2025) almost 40% of current jobs will be replaced by automation. 40% is a scary number. Similar numbers are published by OECD. And not just drivers but also retail industry is endangered: already now self-scanning checkout in supermarkets is available but Amazon plans to extend it and introduce shops that you don’t even need to scan.

And car industry revolution related to automation and AI is not limited to self-driving cars: we are living in the era where more and more countries will be pushing electrical cars, and so much that fuel powered cars might be even forbidden in some countries altogether. It will be interesting to observe how traditional car manufactures will survive electrical car revolution: after all traditional cellphone manufacturer Nokia didn’t survive smartphone revolution: fuel powered cars are like dumb cellphones and electrical cars are like smartphones in more senses than one.

Of course thanks to digitization of all industries and AI revolution new jobs will be created but new skills will be required to get these new jobs. In other words: automation and AI will make many people unemployed but paradoxically companies will be searching for new employees too, however employees with totally new and different qualifications.

Due to many jobs lost to automation and AI there will be big push to introduce Universal Basic Income but also this revolution will lead to increase of gig economy where more and more people will not have permanent job but just short-term contracts or projects. Massive changes in society coming very soon.

Just because many AI software like Machine Learning is not programmed but trained, humans will lose insight into how these virtual vessels, this software, works, and also to predict how it will behave.

In a very long term – probably 2033 at the earliest, sentient AI, general artificial intelligence, is unavoidable too and then singularity is coming and the question rises why still would indefinitely more intelligent AI want to keep humans around? Particularly because humans don’t care what lower life forms like animals think – what is expressed in extinction of many animal species but also in idiotic expressions by some humans who say “like an animal” as derogatory phrase. Just imagine in post-singularity future where AIs are talking to each other and one is saying “of course I didn’t do this that way – like a human – I did it this and this way like an AI that I am”. Singularity means that AI from remote future will be by definition not predictable because their intelligence will exceed human intelligence many times: ant can’t understand you so you will not be able to understand future AI. And who are we kidding: since contemporary AI is using reinforced learning and is practically learning from itself, we can’t even understand how contemporary AI functions – it’s trained or self-trained, it’s not merely programmed by humans.

Finally let’s look ahead millions of years. In science-fiction movies and TV shows there was always belief that in very remote future mankind and all aliens are evolving into beings of light only – pure energy. But else is AI? It’s pure energy. So maybe in remote untrue biological beings will not evolve to biological beings who are just light and energy, but will be uploaded to cyberspace and can merge with AI and thus be projected via holograms as beings made of light and pure energy – non-corporeal beings.

Also: have you heard of Fermi Paradox? Fermi Paradox says that if probability of existence of alien life forms is very high, then where is everybody? Universe is vast and chances are hight that alien civilizations exist and evolved so in each galaxy there should be many civilizations and we should have been visited by now. But no aliens landed in public yet on Earth. Why? Fact that all biological life evolves into AI and purely cybernetic beings might be good explanation of Fermi Paradox: such highly advanced AI civilizations with beings of IQ millions times higher than human are just not interested in us. So future of AI might be future of life on Earth indirectly.

Well, you can see that we do not need to philosophize about remote future to see upcoming AI revolution: spreading of smart speakers, deep learning algorithms and machine learning embedded transparently in many places is already under way.

Future of AI can’t be predicted but one is for sure: massive AI revolution is unavoidable.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s