What it is: Artificial intelligence, or machine learning, has been heralded as the next revolution since the 1960s. Today, it’s here but just unnoticed.
Back in the 1960s when computer scientists first talked about artificial intelligence, everyone thought we’d have super smart computers within a decade. That’s why movies like “2001” featured a completely sentient computer (HAL). Yet for decades, artificial intelligence disappointed people. Every time artificial intelligence came close to matching human intelligence in fields as diverse as chess and medical diagnosis, it always seemed to come up short. Machines could appear intelligent but their biggest drawback was that they couldn’t learn without reprogramming from humans.
That’s all changed with neural networks and machine learning. The basic idea behind machine learning is that the computer can learn on its own. The more data you feed it, the faster the machine learns. That’s why Google’s machine learning efforts found that computers could analyze data on the Internet and learn to recognize cat videos.
Initially, machine learning and artificial intelligence was thought to be something people would be aware of. Yet in the past decade, artificial intelligence and machine learning has quietly crept into society and remained hidden. Perhaps the most prominent case of machine learning comes from Siri.
Back when Apple introduced a natural language recognition system that could understand spoken words, Siri seemed revolutionary. Microsoft quickly introduced Cortana and Google introduced Google Now. Yet while Siri could recognize natural spoken language, its accuracy was spotty at best in understanding. That all changed when Apple adopted machine learning techniques so Siri could improve itself the more people used it.
If you compare Siri now to what Siri used to be like in the beginning, it’s like using a totally different product. Siri is now responsive and accurate in multiple languages. Speak English, French, or Chinese and Siri can recognize your commands. More importantly, Siri is often accurate in understanding your spoken commands. For many people, Siri is just another feature of Apple products, but they don’t realize how much machine learning has gone on in the background to make Siri much better over time.
Of course, machine learning has been used for years in the background to filter out spam, check for credit card fraud, and even identify the best route to take given changing traffic conditions. Machine learning has been part of society for years, but most people never knew it. Even with Siri, most people don’t recognize that machine learning drives Siri. They just think of Siri has another feature where talking to your iPhone now seems normal and not magical or astounding any more.
That’s the future of machine learning. It won’t be a breakthrough that attracts a lot of attention. Instead, machine learning will just quietly work in the background without most people even aware that it exists at all. That means you can expect machine learning to be embedded in future products from self-driving cars to robots.
Machine learning is here to stay. You just may never know it unless you peek under the hood.