Drones are winning the market, and the key...
An insight into Artificial IntelligenceThe term “artificial intelligence” was coined by John McCarthy in 1955. It can refer to the intelligence demonstrated by computers and software or the science of making smart objects and programs. AI often goes beyond ordinary computing. A computer (even a powerful one!) can only handle problems that are well-understood today. Such machines are considered “somewhat intelligent”. Also, AI is not supposed to simulate human intelligence: after all, most of the tasks tech companies want to solve with the help of AI involve a lot more computing than an average human can do.
There are two major branches of artificial intelligence:
- General (or science-driven). Its major goal is to understand how certain things (a language, for example) work and can be replicated by a computer;
- Narrow (or industry-driven). Tech companies use computers and software to solve real-time problems. Need an example? Well, think of embedded systems that operate any smart gadgets out there (like GPS receivers and autonomous vehicles).
Artificial Intelligence: areas of specialization
- Gaming. Remember the famous 1997 match between the world champion Garry Kasparov and Deep Blue, IBM’s chess computer? Although the pivotal move that assured Deep Blue’s victory might’ve been caused by a glitch, the match remains one of the perfect examples of what Artificial Intelligence can do (and the reason why some people fear possible AI takeover!);
- Expert systems. You don’t think computers are capable of making important decisions in real-life situation? Watson disagrees. And we’re not talking about Sherlock Homes’ friend right now. Last year IBM set up a new business center in Cambridge, Massachusetts, to monetize its Watson project. The Watson computer can answer different questions asked in human language. In 2011, Watson stole headlines after its glorious win against Ken Jennings and Brad Rutter, the Jeopardy quiz show champions. This time, IBM wants to apply the machine’s superpowers to healthcare problems. Watson is supposed to analyze all the data generated by smart gadgets & stored in health records, thus enabling doctors to give more accurate diagnoses;
- Voice recognition. Making computers understand natural human languages is definitely not an easy task – and that’s why human translators still earn a decent living. Apple’s Siri is probably the closest we’ve got to it so far. The personal assistant can access any application installed on an iOS device and search through its database. If you want to book a taxi or read an email, you simply say “Hey Siri” and assign the task to the clever program. Siri gets smarter with every iOS update. As of now, it can be integrated into Smart Home systems and manage your new Apple TV. And more is to come;
- Machine learning. Some computer systems can reproduce physical connections that take place in human or animal brain. Microsoft engineers, for example, have recently built a machine that analyses meteorological data and makes accurate forecasts. In the long run, they want to use the algorithm to understand how climate change impacts weather patterns;
- Robotics. Today machines – or robots – can successfully replicate certain elements of intellectual ability and interact socially. The Kismet robot, for example, recognizes voice inflections and human body language & responds to the person he’s talking to.
Examples of Artificial Intelligence
- Vocal IQ. There’s actually more to Apple’s artificial intelligence expansion than Siri. The company has recently purchased Vocal IQ, a UK-based tech startup & the driving force behind the world’s first self-learning dialog API. Apps built with such APIs accumulate previous requests and become more intelligent every time you use them;
- Chloe. Apparently, Siri has a “smarter sister”. Instead of giving algorithm-based answers, the Chloe app interprets your question (where’s the nearest flower shop?) and redirects it to a real person who sends you a personal message. According to Ryan Hoover, Chloe’s “father”, his major goal is to make app users feel like they receive messages from a friend. Chloe can be operated through iMessage or any other messaging app you use:
- Viv. The ambitious project carried out by a tech startup of the same name brings Artificial Intelligence software development to the next level. The company managed to raise $ 30 million in funding and is currently working on a sophisticated voice-controlled personal assistant. Unlike Chloe and other AI programs that mostly dwell on smartphones, Viv will be integrated into every connected gadget out there (from smart thermostats to self-driven vehicles). “Tell Viv what you want and it will orchestrate this massive network of services”, says Dag Kittlaus, the company’s founder and CEO. The program doesn’t have a release date, though Kittlaus hints 2016 will be huge.
Smart Home systemsMark Zuckerberg has recently announced plans to build an AI home automation system for his own house. First, he’s going to explore the technologies that are already available on the market, like Amazon Echo. His next step is to set up voice and face recognition, so that smart thermostats could adjust temperature to the needs (and preferences) of a person who enters the room. He also wants to create a virtual reality solution to visualize project data (and be able to work from home). You think Mark’s not going to live out his Tony Stark dreams? Well, the sci-fi is already happening!
- Works with Nest. The program allows you to connect Phillips light bulbs to the system and make them dim on irregular intervals once you’re on holiday to keep burglars away. Thanks to the smartphone-like touch display and sleek UI, the thermostat is very easy to operate;
- Alexa. Amazon, Zuckerberg’s favorite, went even further. Its Alexa (or Echo) smart cylinder is connected to the Cloud via Wi-Fi. The gadget is shipped along with several microphones; you distribute them over your lodgings, enabling Alexa to catch & interpret commands given in natural language. Amazon Echo is compatible with both iOS and Android, so you can manage the gadget through a mobile application. With Alexa, you can perform multiple tasks – from ordering Chinese food to editing playlists on Spotify.
The use of Artificial Intelligence in software developmentUntil recently AI’s been all about consumer electronics and personal assistants. Now it looks like tech companies shift their focus to application development. It might be too early to judge how helpful smart apps will be, but what we’ve seen so far seems quite promising.
- Wikipedia’s ORES. The world’s largest Web encyclopedia now uses an AI program called Objective Revision Evaluation Service to automatically detect inaccurate content and assign quality scores to articles. The application will certainly help Wiki volunteer editors who have to manually review the website pages 500 thousand times a day;
- RankBrain. Every time your google something, your request may be interpreted by RankBrain, an AI system launched by the company last October. Unlike the standard web search algorithms, RankBrain tries to identify human speech patterns (rather than ordinary keywords!) and deliver more accurate results.
IoT is the future of Artificial IntelligenceIn 2016, there will be 6.4 billion connected devices worldwide (30% up from last year). Approximately 1 billion of these devices will be used by enterprises (guys like you and us who handle payments and store customer data). It’s no wonder the McKinsey Global Institute report raises questions about interoperability, security challenges and failure to use all the data (400 zettabytes by 2018) IoT gadgets generate. Why do we need IoT data in the first place? We can use the sensor readings collected by smart bulbs, fitness trackers and cameras to:
- Establish better communication between self-driven cars;
- Revolutionize the whole Smart Home concept;
- Increase productivity in the workplace;
- Prevent crimes and accidents;
- And do a lot of other cool things!