Artificial Intelligence Moves Beyond Science Fiction

The topic of artificial intelligence (AI) has inspired visionary writers such as Isaac Asimov and Arthur C. Clarke – as well as the creators of “Star Wars,” “Star Trek” and other popular films and TV programs – to produce some truly thought-provoking science fiction.

But as AI emerges from the labs and into our lives in the form of voice-enabled assistants, online navigation and surprisingly well-informed shopping apps, it’s clearly moving beyond the realm of fiction. AI is here to stay, in fact, and Samsung’s Flash memory products are playing an important role in enabling this promising new technology.

An explosion of data

Part of what’s fueling the coming AI revolution is the enormous amount of data that we, collectively, are creating. IDC expects the “global datasphere” – the total amount of data in the world – to grow by about 20 zettabytes (ZB) in 2017. By 2025, IDC expects the datasphere to be increasing by 163 ZB annually, eight times faster than the current rate. (For comparison, the entire World Wide Web contained about ? ZB of data in 2009).

Fortunately, AI technology appears to be well-suited for sifting through our growing mountains of data to discover the valuable insights hidden within.

AI’s adoption is also being driven by the soaring popularity of smartphones, PCs, TVs and other connected devices. Cisco estimates that 19 billion connected devices are now in use, a total that should reach 27 billion in 2021. Much of that growth is expected to come from machine-to-machine (M2M) connections, which are used in AI systems, as well as smart home networks, to exchange information and perform various tasks without manual assistance from humans.

Machines that learn

Today’s most widely used AI applications rely on machine learning, which allows computers to sort through vast amounts of data and “learn” to recognize patterns without being explicitly programmed. Deep learning, a more advanced type of AI, uses multiple layers of data analysis to reveal even more complex patterns and insights.

We currently see at least four main AI market opportunities:

? Image and facial recognition: This type of recognition, which backs up more than a billion photos and videos each day, uses AI to identify familiar faces and places to help users organize and share their images.

? Natural language processing: AI’s ability to analyze spoken language is helping to improve customer service, web searching, translation and other tasks.

? Deep learning: It’s now being used in weather prediction, game theory and genomics – the study of genetic material – where it has helped to drive nearly 20-fold market growth in recent years.

? Autonomous driving: Driverless cars will require ultra-fast, high-volume machine learning to avoid obstacles and adapt to changing routes and traffic. As the auto industry gears up for AI-enabled cars, it’s likely to require 15 times more data center capacity by 2025.

To manage all this new data handling and analysis, we’re going to need more efficient storage architectures, lots of new storage capacity and far more powerful memory chips.

That’s where flash memory technology comes in. Samsung recently announced plans for several new kinds of NAND memory chips that will offer some of the industry’s highest storage capacities and speeds. We also announced new solid-state drives that will set new industry performance benchmarks.

By better managing the enormous influx of new data, I expect these new memory products to enable even more exciting new developments in AI technology. Who knows, they may even inspire a new generation of science fiction.