Decades of research and billions of dollars have poured into developing artificial intelligence, which has crossed over from science fiction to gameshow novelty to the cusp of widespread business applications. Artificial intelligence is an area of computer science where computers are designed to think and operate much like a human brain, supported by advanced forms of computing and software. But it’s only as smart as the amount of information fed into its memory banks. The more information, the smarter it gets. The greatest advancements have been demonstrated in the area of game playing, but AI is now showing its mettle in the business world. “People are starting to kick the tires, looking to see how it can help their business and the bottom line,” said David Schubmehl, who follows the AI field for research firm IDC. “The return on investment evidence is not yet clear, but it’s starting to happen.” Facebook ( FB ), Alphabet ( GOOGL ), IBM ( IBM ) and Nvidia ( NVDA ) are among tech leaders with big artificial intelligence ambitions. “We’re starting to see a lot of companies beginning to use different types of AI tech for various uses,” said Schubmehl. “We’re also seeing a lot of venture capital money flowing in and a lot of acquisitions taking place.” IBM, which may have the deepest AI investments and most far-reaching ability of any company, has pitched “cognitive computing” as a tool for businesses via its cloud-based Watson platform. Facebook is using AI to decipher the best ways to bring Internet service to remote areas of the world and to make its News Feed feature more relevant to users of the social network. Alphabet is using it to enhance Google search abilities, improve voice recognition and to derive more data from images and video. Nvidia has developed chip technology for AI platforms used in autonomous driving features, and to enhance how a driver and car communicate. Alphabet, Mobileye ( MBLY ) and others also are tapping AI in the race for driverless cars. It’s not enough to have Google Maps loaded up: A self-driving system must identify potholes, weather conditions, traffic congestion and other drivers’ behavior — improvising and improving on the fly. AI: Ready, Chess, Go Artificial intelligence, a term coined in 1955, was popularized when IBM’s “Big Blue” became the first computer to beat a reigning world chess champion, Gary Kasparov, in 1997. IBM won again in 2011, when its Watson computer on “Jeopardy!” outsmarted the game show’s two top players ever. Last October, an AI computer beat a three-time European champion in the ancient Asian game of “Go.” AphaGo, built by DeepMind, part of Alphabet’s Google, beat Fan Hui by 5 games to 0, the first time a computer program has ever beaten a professional Go player and a feat thought to be a decade away. On Tuesday, AlphaGo won the first game vs. Lee Sodol, who is ranked No. 5 in the world. Tuesday’s match shows that AlphaGo has made big improvements since beating the No. 633 ranked Fan Hui last year. Google paid $400 million to acquire DeepMind in 2014. But despite rapid advances, artificial intelligence is still in the early stages of business deployment. “A lot of what AI is being used for today only scratches the surface of what can be done,” said Babak Hodjat, co-founder and chief scientist at Sentient Technologies . “It will become so ubiquitous that we won’t even call it AI anymore.” AI Picks Out Your Shoes Sentient emerged from stealth mode in late 2014 with a massively distributed AI platform that companies can use to boost performance. As shoppers at Shoes.com browse through photos of shoes and click on ones they like, Sentient’s technology narrows the selection, so people don’t get overwhelmed by choices. There’s no need for text-based searches or drill-down navigation, said Hodjat. The AI technology deciphers what shoppers are looking for, letting them quickly dive deeper into a catalog to find the perfect item that might otherwise go undiscovered. “We’ve revolutionized the user experience on an e-commerce website,” said Hodjat. “It’s a huge change in the way users interact with products online and therefore drives conversion.” Sentient has received $143 million in venture capital funding, the most of any AI startup, according to research firm CB Insights. Since 2010, AI startups have received $967 million in funding. Intel ( INTC ) alone has invested in 16 AI companies, including Saffron Technology . Saffron says its platform “mimics the fundamental principles of how humans remember and learn.” In a case study on its website, Saffron helped an insurer to identify fraudulent auto insurance claims. Over 10 weeks, Saffron examined 113,000 claims from one year in one state and found three potential fraud rings. It then detected that these rings were part of a larger ring involved with 38 claims, of which the insurance company had paid out about $400,000 in claims. Saffron was able to collect data that identified relevant and unknown relationships, including different providers, demographics and injury descriptions, creating a knowledge store that had never been done before. Saffron forecasts that the insurer can avoid a payout of tens of millions of dollars a year. AI goes by terms such as machine learning and deep learning. IBM calls it cognitive computing. In October it launched Cognitive Business Solutions, with 2,000 consultants skilled in data analytics, cloud computing and other areas. More than 500 companies have deals to use Watson, as a cloud service, to develop commercial products, apps and services. Turner Broadcasting on Feb. 29 signed a deal to use Watson in its ad sales efforts. The Time Warner ( TWX ), unit owns TBS, TNT and CNN. Using IBM Watson, Turner expects to parse through all manner of data to help draw in more advertisers and provide them with greater impact. Watson Can Predict Low Blood Sugar Watson Health is a platform for physicians, researchers, insurers and other companies focused on health and wellness. Medtronic ( MDT ) is collaborating with IBM on personalized care for people with diabetes. By analyzing patient info and data from Medtronic devices, Watson can predict low blood sugar programs three hours in advance. Artificial intelligence is all about using advanced technologies to help develop brainy reasoning from disorganized information — unstructured data — to derive accurate decisions. “What we’ve done with all our research is to really understand how to add unstructured data to a decision,” said David Kenny, general manager of IBM Watson. Unstructured data comes from a multitude of sources that is not organized. It can be the data bits from photos, medical images or video and audio transmissions. It can be the tons of data that flows in from cameras or sensors in buildings or smartphones, from Web traffic, tweets, inside emails, government filings or business documents. The more information an artificially intelligent computer can digest the smarter it gets. Today The Weather, Tomorrow… Collecting data on a massive scale is among the reasons IBM acquired several Weather Co. properties, including Weather.com, Weather Underground and mobile and cloud-based assets, for a reported $2 billion in October. IBM will be able to analyze data from more than 2 billion weather reference points, over 40 million smartphones and 50,000 airplane flights per day, letting it offer a broad range of data-driven products and services to more than 5,000 clients in the media, aviation, energy, insurance and government industries. IBM said it can provide predictive weather analytics along with real-time analysis of social media chatter, detailed understanding of transportation flows and other data that will benefit retailers and distributors. The Weather Co. assets will serve as the foundation of a new Watson unit focused on the Internet of Things (IoT). Early this year, IBM announced that it would invest $3 billion to build out IoT products and services. “The more we do this, the smarter Watson gets and the smarter AI gets,” said Kenny. As to why AI is becoming more widespread: “It works,” Kenny said. “Users of AI are saving time and money. They’re making faster decisions and getting better outcomes.” An analysis of Facebook and Alphabet’s Google by research firm Innography shows a surge in AI patent filings that began in 2010. Alphabet currently has more than 3,000 AI patents that are active or pending government approval. Facebook has about 870. AI Knows Where You Live Facebook last June opened an artificial intelligence lab in Paris with a goal of improving the way users interact on the world’s largest social network. “It’s our hope that this research will ultimately help us make services like News Feed, photos and search even better and enable an entirely new set of ways to connect and share,” Facebook said at the time. It also has AI research teams at its Menlo Park, Calif., headquarters and in New York. Facebook in February revealed its Connectivity Lab project. It used AI to analyze 8.3 million square miles of land, using roads, schools and other structures to determine where and how many people live in a given area. That can tell Facebook’s Aquila drones where to go to provide Internet access to less-developed areas, as part of the company’s Internet.org initiative. “This data will give us a greater understanding of how populations are dispersed, so governments and others can prioritize investments in infrastructure, from transportation to healthcare and education,” Facebook said.