Okay, let me break down the acronyms for you: – Artificial Intelligence (AI), Machine Learning (ML) and Deep Learning (DL). Well I was recently introduced to a new blog site and thought one of the articles posted there was quite interesting so I have decided to share it with you, let me know what you think.
There now exists a general consensus that Artificial intelligence will shape our future more powerfully than any other innovation this century.
Rewind to around about half a century ago, 1950 precisely, Alan Turingconducted a test of a machine’s ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human.
In his test, he actually deemed a computer to be passing the test of intelligence if it can fool a human interrogator. The test was simple in nature, an interrogator interrogates two separate entities 1. a human and 2. an AI system but cannot see them.
If the interrogator cannot tell apart the two as in which is the human and which is the AI agent then it says that it fooled the human interrogator and it showed to exhibit some form of intelligence because it was able to do that.
In his paper, Alan Turing also suggested major components of AI including knowledge, reasoning, language understanding and learning. While this was a good test, Turing inadvertently exposed AI to be completely unlike humans, well OK maybe like some humans and that is essentially the concluded deception.
Turing’s conclusion arguably points to the humanoid interface of machines to be quite flawed in instances where an AI is on board humanoid robot hardware. Therefore since AI has the capability to outsmart humans, it might be more worthwhile as a tool rather than any other potential applications. Similar to electricity for instance, intangible yet effective and far away from the humanoid premise because to be human is to be vulnerable, AI on the other hand, AI is hardly vulnerable, in fact it is immortal, and as proved by Turing, capable of undetectable Deception of humans.
Ever since, we have been developing humanoid robots with human like attributes for a long time now but is this trajectory of robotic projects the right direction? Perhaps if robotics projects focused on tooling for potential applications, robots may have far more reaching use cases than the common humanoid premise. It maybe that the weaponry potential of this approach renders it a requirement but we already have autonomous vehicles and unmanned predator drones.
It is a question of why have a robot to operate a machine when the machine itself can actually self operate? For example would you rather build a robot and equip it with a gun or would you give the gun mobility capabilities? Autonomous vehicles are a very good example of tooling that has been achieved by AI. Do we really need robots at all? This however is not the big question.
Since the recent rise of AI, following on from about four decades of false hope, researchers have conducted several experiments in which AI has proved to have much more superior intelligence than a human being, Supper-intelligence. Intelligence surpassing that of the brightest and most gifted human minds.
University of Oxford philosopher Nick Bostrom defines super-intelligence as “any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest”
An intelligent agent can potentially out think a human being in any area of cognitive performance. AI has since demonstrated this in recent times thanks to rapid advances in data storage and computer processing power.
In 2016, DeepMind, AlphaGo AI, defeated one of the best human players at Go. AlphaGo’s first formal match was against the reigning 3-times European Champion, Mr Fan Hui, in October 2015. Its 5-0 win was the first ever against a Go professional . AlphaGo then went on to compete against legendary player Mr Lee Sedol, winner of 18 world titles and widely considered to be the greatest player of the past decade. AlphaGo’s 4-1 victory in Seoul, South Korea, was watched by over 200 million people worldwide. AlphaGo was able to imitate other players, evaluate a board position and determine the probability of winning from that position and look through different variations of the game and attempt to figure out what will happen in the future. Mr Lee Sedol, can be credited for that one win staked against such computing power.
In 2017, comes Libratus, the Superhuman AI for No-Limit Poker. Heads-up No-limit Texas Hold’em is the most popular variant of poker in the world. It has become the main benchmark challenge problem in game theory for AI in imperfect-information games. Unlike Chess and Go games which are perfect-information games were both players know the exact state of the game at every point, Poker is an imperfect-information game. Part of the state is hidden from a player because the other player has private information. The imperfect information scenario is common in the real world which makes such a game suitable for modelling real world strategic interactions. In January 2017 Libratus beat a team of four top-10 headsup no-limit specialist professionals in a 120,000-hand 20-day Brains vs. AI challenge match. That is the first time an AI has beaten top humans in this game. Libratus beat the humans by a large margin (147 mbb/hand), with 99.98% statistical significance. It also beat each of the humans individually.
We are already surrounded by machine intelligence, it is all real now and fundamentally changing our lives. We are officially dependent upon machines. We can talk to our phones and they understand. Machines can see, speak and listen. We have self driving cars that promise to eliminate car accidents. Machines can save our lives in medical applications. Thanks to computing we are starting to understand our own genetic code, information that could be used to personalise treatments. The one sure common benefit that this all seems to have is the extension of our own intelligence. A smarter machine is a better machine. Today we have all already used AI in one way or the other. If you have run a Google search you have already interacted with AI at some level. Computers today run machine learning algorithms to draw patterns and make some accurate conclusions.
Cambridge Analytica involved in the recent personal data scandal used AI for the computation of users data. AI, by reading the user data got a good glimpse into behaviours and maybe even make associations of those people with other internet interactions they do or have done, enough to respond in a way that can influence or sway their choices in a direction it so chooses or is required of it.
AI is achieving these feats through Deeplearning, the concept of getting machines to learn by themselves, by reading data and figuring things out, and becoming smarter and smarter. There is already significant talk about displacement of jobs. With routine work already getting automated. Jobs will be done faster, cheaper and without any breaks by machines. Drivers will be the most quickly affected within the next three to five years. These changes are likely to provoke a much greater divide between the haves and have nots. Students enrolling into universities to follow careers such as accounting, doctors, radiology for instance might find it difficult to get a job in those fields in the future.
The Davinci Robot is used today in open surgery. It is not the future, it is the present. With AI on board the Davinci robot will be able to learn more from each surgery it is applied and by understanding what works and what does not over time the robot will be able to perform surgery by itself or under human supervision. Cases of hysterectomies are increasingly being performed robotically. A surgeon who used to perform 150 such cases per year, now has this reduced to 1 thanks to the Davinci robot. This renders the surgeon to no longer remember how to open patients anymore due to lack of consistent practice.
The machines getting better than us at exponential rates makes a compelling case for our need to stay ahead of the machines. Today, computers are learning to read emotion. Researchers have been working to give computers the capacity to read our feelings and react, in ways that have come to seem startlingly human. Our faces are organs of emotional communication. It is estimated that we transmit more data with our expressions than with what we say, and a few pioneers dedicated to decoding this information have made tremendous progress. Perhaps the most successful is an Egyptian scientist living near Boston, Rana el Kaliouby. Her company, Affectiva, formed in 2009, has been ranked by the business press as one of the country’s fastest-growing startups.
These significant demonstrations of super intelligence capabilities of AI will leave us not much choice than to actually begin to consider AI on board ourselves otherwise the age of machines will render us obsolete as a species. We are swimming in a sea of data and AI is the smart agent which can quickly make sense of it faster than previous computation capabilities could achieve. It is either we acquire this ability ourselves or we leave it in the hands of machines and computation, that we do not know whether to trust or not. It is in this assertion, where lies the big question.