#I love the idea of Reader meeting ‘bots in vehicle form first so they can properly freak out later at how tall they are in bipedal mode

[delicateartisantrash]

W.I.P. Soundwave & Knockout

When you can’t decide which ‘Con to hang out with, so both end up folding themselves into a tiny two car garage.

#goals?

[grimurfjeldsted-blog]

Quick Guide to Digital Tech - Artificial Intelligence

Introduction

This is the first article in a series "quick guide to digital tech", where I will cover some of the most discussed topics in digital technology. With the series I aim to give a simple understandable overview, with examples most readers can relate to or have heard of. In the end, I will give a short comment on where I think the technology is today and where I see it heading in a near and far future.

Part one: Artificial Intelligence (AI).

Overview

Artificial Intelligence may be the most disruptive technology the world has seen since the Industrial Revolution. It describes every aspect of learning or any other feature of intelligence that can be described so precisely that a machine can simulate it. A machines ability to solve a problem, usually solved by human beings. These can range from simulating complex functions of the human brain, programming a system to use general language, determine and measure problem complexity, as well as self-improvement and abstraction. We even see AI slowly moving into creative areas like algorithm-driven design, which most certainly is discussed in design schools these days.

Most AI is so-called weak AI. It is not really weak and can hold powerful algorithms, but is called weak because of its focus on one narrow task (sometimes also called "narrow AI"). Deep Blue by IBM, the first chess-playing computer who beat a reigning world champion - Garry Kasparov was a narrow AI. It solved a certain problem, in certain conditions under certain rules. Even if the accomplishment is not to be underestimated especially not for 1996 standards, it did only one thing really well - playing chess.

Strong AI on the other hand aims to solve a matter of general intelligence, a far more complex area. It can try to simulate consciousness, sentience and mind. Strong AI has yet to be developed, but several projects aim to do so. IBM's Watson, with its cognitive capabilities and Googles DeepMind (see below) might be ranging somewhere in between the two terms.

Machine learning is a type of AI that provides machines with the ability to learn without being explicitly programmed. E.g. a robot that usually is programmed to perform repeatable tasks, uses machine learning to adapt to changing circumstances. Deep learning is a subset of machine learning that, in very simple terms, can be thought of as the automation of predictive analytics.

Examples

Some of the well known examples of AI are best described, when we broadly define and categorise the term intelligence.

General intelligence - learning that enables the learner to be able to perform better in situations not previously encountered.

Reasoning - to draw conclusions based on the situation at hand.

Problem solving - find X, if given a certain starting point or challenge.

Perception - analysing a scanned environment, features and relationships between objects.

Language understanding - by following syntax and similar rules used by humans, or dolphins for that matter.

Let us look at some common examples of AI use in the context of the categorisation.

General intelligence:

DeepMind: acquired by Google in 2014, DeepMinds mission is to "solve intelligence" by simulating the human brain. Differentiating from Narrow AI's, DeepMind is on a mission to program machines with the ability to learn for themselves, regardless of context. Further information on DeepMind [http://www.wired.co.uk/article/deepmind]. It is worth the read.

Reasoning:

Amazon: Amazon's transaction AI has been in place for quite some time and we know it from the recommendations we receive on the platform. Its algorithms refine the shoppers data, like searching, browsing and shopping behaviour and allows the platform to predict and recommend our next purchase. Amazon is taking this even a step further in their anticipatory shopping project. It is supposed to anticipate a shoppers needs, and send you items before you actually need them. It is not quite there yet, but is certainly on the horizon.

Netflix: provides highly accurate predictions and recommendations based on users interaction and reaction, by matching the data with its entire database of movies and series. A feature we know, hate or love from our favourite music streaming service.

Computer games: If your neighbour like mine had a Commodore 64 in the early eighties, you would probably have encountered a simple narrow AI for the first time. AI development in the digital game industry has increased as massively as the industry itself has grown.

Problem solving:

Deep Blue by IBM (see above) and AlphaGo: AlphaGo is an application by the sophisticated DeepMind by Google developed to beat a human player in the even more complex game of "Go". I highly recommend the bio of Demis Hassabis, founder of DeepMind by the Financial Times if you need inspiration for dinner table conversations [https://goo.gl/DqUt45].

Mastercard: uses AI technology for fraud detection to increase real-time approvals for transactions and reduce false declines. The system leverages machine learning by examining and analysing patterns in normal vs. abnormal spending behaviours. [https://goo.gl/T8GD5g]

Healthcare startup AiCure: is using mobile technology and facial recognition technologies to for remote patient monitoring to determine if a patient is taking the right medications at the right time. [https://aicure.com/]

Perception:

Self-driving cars: are required to understand complex traffic scenarios and moving objects, that the vehicle can recognise and react to. AI helps anticipating intentions of traffic participants, and interpret the traffic scenes ahead. It is a huge challenge in AI development to comprehend such demanding contexts, but will eventually make traffic safer and more efficient. What we see in the street of the Bay Area these days, can be seen as AI's driving lessons, training before it can get its license.

Smart home: most of us are familiar with Google Nest, a learning thermostat, that uses AI and behavioural algorithms to predictively learn and adjust peoples heating preferences based on personal needs. It also integrates with other smart devices like the Nest Protect, a smoke alarm that will communicate with the thermostat in case of a fire. Also a simple example of the "social internet of things", a topic covered in the next article of this series. [https://nest.com/]

Language understanding:

Virtual assistants: Apple's Siri, Microsoft Cortana, Google Now, Amazon Alexa, etc. help you find useful information via voice input. The AI collects information from users, processes it to better recognise certain users speech and serves results based on preferences and behaviour. Virtual assistants will likely develop the ability to anticipate users's needs like the one Amazon is experimenting with in their anticipatory shopping project. Another take from Amazon is the Echo speaker, that besides serving information based on voice input, lets you control elements of your smart home, and of course allows you to order products (as prime members - of course).

Chat-bots are another example of virtual assistants, ranging from very simple often artificial feedback to complex systems used in robotics, that can interpret body language and facial expressions in order to adapt the robot's form of interaction with a person. Check out Amy, an email-bot for scheduling meetings, that took the writer of this article four emails to figure out, that he was wishing an AI bot a great weekend with its family. [https://x.ai/]

State of the technology and future outlook

Are the machines taking over? Not yet at least. However AI is making its way into our lives, affecting how we live, work and entertain ourselves. The futuristic, western-themed TV series Westworld from HBO that aired in 2016 gives us an outlook, of what AI might look like in the future. Its popularity taps into our fascination of the technology.

While AI today still is in its infancy, innovations in processing power, sensor technologies, robotics, cloud computing and advanced algorithms, will evolve AI with immense speed. That in the context of exponential growth and the lack we as humans tend to have for grasping ideas, we at this point in time can not built - the future outlook for AI is hard to predict. Artificial intelligence that does not need programming, will learn and experience by itself will rapidly become more smart than humans, whether we like it or not.

Apart from the discussion if robots and artificial intelligence will take or create jobs, you will have encountered the last couple of months, what is sure is the huge change it will bring to society and the way we live on this planet. If it is a good one, fortunately still lies in human hands - for now!