This Week in Analytics & AI News
You’ve heard the terms, but they bear repeating: blockchain, augmented analytics, and persistent memory top the list of trends shaping big data’s future. One AI leader makes the case for making artificial intelligence a lot more boring. The banking industry agrees, but not in the way you’re thinking. Intel has put together a laundry list of recommendations for getting the United States’ AI strategy back on track. And a new supercomputer at the University of Cambridge “rains” supreme.
The Next Wave of Big Data and Analytics
In the short-lived satirical Showtime series “Happyish,” Swedish thought leader Gottfried constantly repeats the line, “Business today moves at the speed of ideas. You don’t have to like it, but you can’t ignore it.” While it’s meant to be tongue-in-cheek, there is some truth to it. Gartner knows that. Enterprises know that. And in the world of big data analytics, you simply can’t ignore the top 10 trends—like explainable AI and continuous intelligence—that will disrupt the next several years.
Make AI Boring: The Road From Experimental to Practical
We’re quickly reaching the point where AI is baked into everything—your phone, your house, everything. So saying these things run on AI is like saying our bodies run on these revolutionary things called “nutrients.” Hilary Mason, GM of Machine Learning at Cloudera, says it’s time to stop hyping AI’s wow factor and instead focus on the outcomes. As she so wisely puts it, “Let’s make AI boring—practical, repeatable, and scalable—to drive real business results.”
Banking Chatbots Need More "Advisor" and Less "Robo"
But boring doesn’t have to be robotic. Banks are feeling the heat more than ever to imbue their digital assistants with the language and tone that mimics an actual human being. That means real live UX developers with real hearts and real emotions should focus on writing algorithms that reflect the bank’s narrative while portraying a real sense of warmth and empathy. Talk about dollars and sense.
Intel Gets Specific on a National Strategy for AI
It’s no secret that governments all over the world see AI as a catalyst for economic growth and a means of improving their citizens’ lives. The U.S. is no exception, but in many ways it’s falling behind. To urge government officials in the right direction, Intel has put together a 4-point plan recommending the steps the country must take to help its citizens embrace the benefits of AI technology, bring those benefits to life sooner, and move forward with an AI strategy of its own.
Architecting for AI Workloads
Over the years, we’ve exhausted every cloud metaphor under the sun (see, we just can’t help it). Some are forced, others are pure comedy gold. But the University of Cambridge just dropped the mic in our face with a supercomputer called “Cumulus.” You may have heard about the data explosion. You may have also heard that AI, analytics, and HPC are converging. Just like a cumulus cloud, the Intel-powered system serves as a single HPC cluster to shower researchers with support for data analytics, machine learning, and large-scale data processing. Well played, Cambridge.