AI is the New Electricity

Brain Cells and Deep Space

People have always been fascinated with mysteries of intelligent robots. So far this has been limited to myths and legends. Form the legend of Golem to HAL 9000 in 1969 film Space Odyssey, our imagination has been fuelled by some unknown intelligence that will take over our lives. Today, artificial intelligence (AI) is entering a stage of being – to paraphrase the author of the mentioned movie, Arthur C. Clark, indistinguishable  from magic.  The impact of AI is going to be huge in the coming years. In a conference in May 2016, Andrew Ng, Chief Scientist at Baidu and one of leading researcher into AI stated: “AI is the New Electricity.” We are seeing the beginning of a shift to a world where software will dominate and control our lives.

The idea of intelligent machines is synonymous with computers. The first computes in the 50s and 60s were called “electronic brains”. Ironically they ware far from intelligent but basically good a calculating both fast and accurate. Despite the consensus  that these machines possessed some form of intelligence it quickly became apparent that machines were good at doing things that humans are not so good at, at least very slow on the average. Calculating 1,000 five digit numbers is both tedious and slow for humans. The risk of mistakes is also pretty high. For computers this is straightforward, fast and accurate. However, it turned out that tasks that humans find easy, such as understanding language or recognising objects in a picture is notoriously hard to program a computer to do.

Artificial Intelligence started as a field sixty years ago in 1956 summer workshop at Dartmouth Collage in the USA. The workshop was organised by John McCarthy, and attended by Marvin Minsky, Claude Shannon, Nathaniel Rochester and others that would become very influential on the field in the decades that followed. The goal of the workshop was to “solve kinds of problems now reserved for humans…if a carefully selected group of scientists work on it together for a summer”. That proved to be embarrassingly too optimistic.

The history of AI is full of “springs” – new hope for new ideas, and “winters” when people realise they hope they had was limited or simply did not work. The general conception was been that AI has never been able to deliver its promise. However, many of the advances in computer science is due to research in AI. As soon as something became practical and worked, for example new way of searching though wast amount of possibilities, it become known as something else. Some ideas just did not work due to the limited capacities of the computers at the time. For example, the ideas of building a computer system that was similar to the brain using the idea of neurons and connection between them, came as early as the 1950s. Some mathematical work was even done before the first computers. However, the computers of the 1950s and 1960 were simply not powerful enough to be able to achieve any success.

The first true public success of AI came in 1996 match between IBM DeepBlue and chess master Gary Kasparov. People realised that machines could become better than people in some cognitive tasks. In 2011, AI hit another milestone when IBM’s Watson supercomputer won the television quiz show Jeopardy. Pitted against the two most successful players the AI managed to win. The game requires understanding of language so this signalled a new era in natural language processing.

In 2012, Google posted a seemingly uninteresting blog labeled “Using large-scale brain simulations for machine learning and A.I.”. In the post, Google explains how they built a neural network, a form of machine learning or deep learning, had discovered how to recognise cats in Youtube videos. If there is anything in abundance in this world it is Youtube videos of cats.

So how does this work? We know how traditional programming works. You write programs, series of commands such as expressions, variable assignments, if-statements and while loops and so on. These instruction tell the computer what to do and the computer will execute the commands. If there is an error or a “bug” you edit your program and run it again. Neural networks are not like this. They are of course programs but instead of programming the task, like finding cats or understanding language, we build a neural network or “brain” and train the network to learn how to do its task. For example, Google’s DeepMind, created an artificial intelligence program using deep learning to play Atari games. The only input the program was how to control the game (for example, move a bar left or right) and that high score should be as high as possible. The program then trained to master the game.

The cat discovery was the beginning of a new AI spring. And of course those who have been following AI research for a long time, like myself, took this with the usual skepticism, sort of “here we go again” attitude. Neural networks did not work in the past, why would they work now?

Three things are now different. First, machine learning algorithms have improved over the years. Many academic papers are published every year and the knowledge increases. Quick search on Google Scholar revealed 638.000 hits dated since 2012. Second, vast resources in computing power where you can build 20.000 GPU (Graphical Processing Units) computer cluster. This is far from the computers in the 1960s. Thirdly, the huge data available to train AI networks. The amount of data generated each day – Big Data, both by people and devices is input for machine learning.

In just the last few years, there has been an explosion in AI solutions coming the market. In most cases this is not obvious since AI, just like electricity, will not be a product but an enhancement to our lives. Just like people wanted light in their houses, not electricity for its own sake, people want the products that AI will bring. It will come in hidden form, making the tools we use more clever and convenient. In few years our personal digital assistant will be something we cannot live without.

This text is based on a new addition to the 2017 edition of my textbook, New Technology 

 

When Things Start to Connect

smart city and wireless communication network abstract image visual internet of things

We take a lot of things for granted. We use hundreds of objects everyday without thinking about them. Clothes, coffee machines and coffee cups, cars, roads, traffic lights, smartphones, computers, showers, newspapers, books, chairs and so on. But what if these objects were not so “dead”? Rather they could respond to their environment and we could communicate with them. Its not clear if we want to communicate will all objects, but for some it might be useful. We could tell our TV to switch on to a particular channel. We could tell the coffee machine to brew a double expresso. The car would inform us that it is due for a checkup and could even suggest a time slot using our calendar. This is what we call the Internet of Things and we are now in the early stages of this new technical wave.

Taking this to an enterprise level, consider products in a store. What if they could be tagged with a radio ID so that they could identify themselves? This is exactly what Radio Frequency ID or RFID does. For years, these small tags have been put on products, shipping containers, livestock and so on. With these tags, systems are able to track them. But an ID is kind of simple. We could also put sensors all around a city to monitor all sorts of changes in the environment. The enabling factor for this is that computers, sensors, wireless capability and so on are getting so small and so affordable that we can enhance normal objects that we use with “smartness” and connectivity.

So what exactly is the Internet of things? Professor Sanjay Sarma of MIT, offers a small but insightful experiment. Ask a kid about the lights in the house. Point to the lightbulb in the ceiling and point to the light switch on the wall and ask them “How is it when I hit the switch, the light bulb comes on?” Now if you think about this, us the older folks, kind of take this for granted. We know there are wires in the wall and ceiling connecting the bulb and the switch, and the switch will just make the connection and activate the electric current. But what would the kid say? The obvious answer is “Wi-Fi.” The light switch will of course talk with the bulb and ask it to turn the lights on! And this is what Internet of things means, having everyday objects connect to the Internet and send and receive data and commands.

While the question of the lightbulb and the switch may be a cute experiment, the eery question remains, which approach is really more cleaver.  If we think about it, the way we are building houses has remained basically the same for decades including lighting. The incandescent light bulb still used today is basically the same as it was when invented by Edison. Wi-Fi connected light system already exist today. Philips, the Dutch technology company, released in 2012 Hue, a Wi-Fi enabled lighting system for homes. You connect a base station (which is called a gateway) to your Internet router. Then each Hue lightbulb (which is called things) will wirelessly talk with the base station. With this setup you download an app to your smartphone and use that to send commands to the lightbulbs via the base station. Since each lightbulb can display any combination of colour the possibilities for lighting the home become endless.

While multi-colour lighting is really cool thing to have – and it is, there are more practical approaches to Internet of things. To explain the real benefits of this technology and why it will have huge impact as it transforms the world, consider an agriculture thing, a tractor.

Efforts to use machines for agriculture started in the 19th century with steam powered vehicles and in the first part of the 20th century the use of tractors and other farm vehicles and replaced manual labour on farms in the developed world. Tractor is a thing. It is just an object that does its job controlled by the operator. It can be used to tow equipment for ploughing, harrowing, planting and so on. But basically, machines like tractors are just analogue machines with the same functionally for decades. What if we take this tractor thing and add to it computing capabilities and sensors? This is the example Harvard professor Michael E. Porter and PTC CEO James E. Heppelmann use in their landmark Harvard Business Review article How Smart, Connected Products Are Transforming Competition.

We equip the tractor with small multiple computers. These could be location sensors, temperature and humidity sensors, and even cameras. Now we have a smart thing, a smart tractor. However, a smart tractor may have new and powerful capabilities but what did this smartness really add? It could provide useful information to the driver about the land and positioning in the field.

iotexplained

An isolated tractor may be smart all right, but still it is limited as the information gathered by the sensors are just stored in the tractor and perhaps just displayed there. To make the tractor more useful, we add wireless capabilities. With that we have a smart connected thing or a smart connected tractor. But still, the real usefulness comes when we take all the information from the sensors and, through the wireless network, use them to create an ecosystem – a farm system.

This is where the true benefits of Internet of things is realised. All this information can be analysed and visualised allowing the farmer to make better decisions about how to manage the harvest. In addition to this, the field itself can have multiple sensors to measure all sorts of things and send these back to the farm system. With a system like this the field can provide valuable information in real-time, allowing the farmer to respond.

Agriculture is a perfect example of how Internet of Things can transform an industry. While there have been many advances with better equipment in the last decades, farming is still very reactive and built on imprecise knowledge of the numerous variables that must be adjusted on a daily basis in order to optimise the production and yield of crops and products. With a farm system, farmers have much better oversight of their farm, can use the land in a much more efficient way and become more productive in a cost effective way.

Farming is just one example. Now think about healthcare, transportation of goods and public transport, urban planning, retail stores, factories, and the list goes on. The Internet of Things is going to transform the world, both businesses and our lives.

This text is based on a new addition to the 2017 edition of New Technology 

Through the VR Desert

slushplay

The year 2016 is the year VR takes off. This is the year where several high-end VR headset, such as Oculus and HTC Vive are released. And this is the year of the VR startups and VR hype. However, it might also be the year of VR setback and disappointments. Indicators might be that the headsets are barely ready and they are far from mainstream. There simply are not that many people that have VR headsets. And even if they wanted, they would need a PC that is able to power the experience. We still have 3-5 years until we have wide enough adoption of this technology and the immersion and quality needed. Until then it is a walk in the desert.

The second Slush Play conference was held in Reykjavík during 28-30. September 2016. The theme var VR and gaming. The conference is small but getting strong traction. In his opening remarks, Hilmar Veigar from CCP, an early player in VR, explained the company move to VR. Hilmar set the tone of the conference. VR works. It is here. However, we are early and we still need to figure things out. We need to get better headsets, smaller and cheaper. Research is still being done in academia to understand some of the subtle things of tricking the brain in a believable and immersive way.  So it will take 3 to 5 years to get to true VR experience with full immersion and to reach a market that can support big production games. Until then, Hilmar suggested, it would be a walk in the desert.

However, a walk in the desert has its opportunities. It is a quest that few will take but the brave ones have a chance of becoming leaders and creators in a new industry. VR is a new platform, just as the smartphone, the web and the PC. Platforms like these follow the classic S-curve. In their beginnings they are just a dream of crazy visionaries. Then when all the enabling technologies are ready, the new platform begins to emerge and work. The performance of the platform starts to increase as it gets better and more people start to use it. We then get into exponential growth as the performance increases rapidly. Finally, it will level off thus forming a forward slanting S. As we saw with the new released iPhone 7, it has reached the top of the S-curve, just as the PC has and the Internet.
waves

Platforms such as VR usually develop in two waves. First wave is overhyped with lots of expectations and abundance of financial capital. But the solutions offered are similar to solutions older platforms, as people, used to the old platform (PC, smartphones) are taking old way of doing this and cramming them into the new platform. For example, the first web pages looked like flyers and brochures or research papers. Then in the second wave, people figure out what the platform can do. For the web we saw the Web 2.0 wave after the dot-com burst. Second wave is much more sober with realistic expectations and production capital.

So we are in the early stages of the VR platform wave and the smartphone wave is ending. In the coming years we will see VR headsets get smaller and lighter and at the same time better and cheaper. As the components needed to build those improve so will the possibilities. Understanding of how to create immersion will increase. When the second wave comes the desert walk will pay off.