Self-driving cars, voice-assisted speakers, the navigation and health apps you use – they all have Artificial Intelligence (AI) in common. Vaibhav Sharma tells us about this futuristic technology
Picture yourself cruising down a flyover without your hands on the steering of the car. Your car spots a stalled truck, smoothly shifts lanes and carries on. Or you’re worried about a health issue, but it’s such a hectic time at work that you can’t visit the doctor. So you take out your phone, fire up a diagnostic app and click a photo. The results arrive within minutes on your phone. Phew! There’s nothing much to worry about.
That’s what Artificial Intelligence (AI) can do for you. In the last two to three years, AI has developed more than it has in the last decade. Machines, applications and entire systems that can ‘think’ and process information faster than regular computers, will change how we live and play by offering powerful, laboursaving solutions to everyday issues.
AI in our lives
At the most basic level, any technology that can perform learning and problem-solving tasks that would traditionally require human intelligence can be termed as AI. This involves two parts: loads of data and the algorithms used to crunch it. While today’s AI is more accurately described as ‘machine learning’, we’re on the way to a magical future where machines might even be able to ‘think’. But even the current fledgling AI systems process data faster and with greater efficiency than we would have imagined a few years ago. And, they learn from their mistakes.
Your personal devices offer the closest AI experience of all. Many smartphones now include processors specifically designed for AI crunching. Voice assistants are de rigueur on every operating system. In the future, they will get smarter and offer better speech recognition and output, making a natural conversation possible. Cameras are also benefiting from AI, with many devices using algorithms and databases of photographs to suggest what setting and even framing to use. The Google Clips smart camera operates continuously, using its AI technology to save photos it thinks are good, while Palo Alto-based startup, Lighthouse, offers a smart home security camera that uses 3D sensing to detect familiar people.
Self-driving cars involve the collection of data (either in real-time or in advance) to create a map that the car can follow. That’s where AI technology comes in. It speeds up processing so that vehicles can react promptly to changing road conditions. For example, consider what a self-driving car ‘sees’ (through cameras, radar and, perhaps, lidar, a detection system that works on the principle of radar): everything from road signs and slow-moving trucks to jaywalking pedestrians and cars changing lanes without warning. That’s right – AI has to not just identify what an object might be, but also compute the appropriate course of action. Think about it: if your car couldn’t tell the difference between a road sign and a pedestrian in the dark, it might come to a halt each time it spotted a sign, or more worryingly, ignore a pedestrian about to step on to the road.
Another example is a scenario where we reach a time when manual and self-driving vehicles share the road.In such a case, the latter could use AI to anticipate a few of the dangerous things humans do on the road such as abruptly changing lanes.
Meanwhile, scientists are of the opinion that the first generation of self-driving cars will use AI for other purposes as well: to inform users what the technology is doing in order