Big50 2017 Tech Predictions – Part 2: Machine Learning and Automation

Machine Learning, Automation, and Smart Devices Three Hot Trends to Watch in 2017

Yesterday, we looked at predictions about how AI will remake datacenter operations, spark M&A activity, and even alter the general fabric of everyday life.

In today’s predictions, AI maintains a starring role, with startups predicting the rise of assistant-based automation, AI-based security, and even automated BI for businesses both large and small.

Here are three more big predictions from Big50 startups:

1. Machine Learning Unlocks Business Intelligence

A big problem with the Business Intelligence (BI) space has been how thoroughly information gets siloed within organizations. Data tends to get locked in specific applications accessible only by a handful of select employees in a single department. Then, it’s nearly impossible to pull that data into BI software for horizontal use across the company.

Dan Udoutch, CEO, of the analytics and machine learning startup Alpine Data, believes that problem could get closer to being solved in 2017. “When we effectively scale machine learning, we can greatly increase the ability of an enterprise to take action on data. Instead of presenting a small number of business users in the enterprise with historical statistics à la business intelligence, companies can bring specific recommendations to thousands of front-line individuals responsible for taking action on behalf of the business.”

This is a major goal of predictive analytics, but until recently, it seemed completely out of reach.

Yet, Udoutch believes an even bigger trend could be afoot, something that broad-based BI is only a small part of.

“We’re on the cusp of [a reality] where our everyday work apps and devices shift from repositories to assistants — and we need to start planning for it,” he cautions.

Today, much of what employees do every day is dictated by the apps they work with. And what you do in those apps is already partially pre-determined by menus and features.

Of those app-related tasks, how many involve workers really relying on their brainpower and how many are just mundane, routine, boring chores?

“[Workers] log into an app, go through a checklist, generate a BI report, etc. In contrast, AI could automatically serve up 50 percent, or more, of what a specific employee needs to focus on that day, and deliver those tasks via a Slack app or Salesforce Chatter. Success will be found in making AI pervasive across apps and operations and in its ability to impact people’s work behavior and tasks in order to achieve larger business objectives,” Udoutch predicts.

2. AI Plugs “Human Middleware Security Gap”

We tend to say this every year, but with the Yahoo! breach, the theft and auction of NSA hacking tools, and Russian-backed cyber-meddling into the U.S. election, 2016 was the year that our cyber-security vulnerabilities really hit home.

A problem businesses have when trying to ward off hackers and other bad actors mirrors what counter-terrorism professionals struggle with: your cyber-security defenses need to get everything right all the time, but the bad guys only need to find a single weakness to exploit once.

Unfortunately, all too often, people are that weakness.

For instance, too many security practices are still manual. Implementing cyber-security protections, for starters, is exceedingly labor intensive. Thus, human error inevitably leads to gaps in security. We’ve seen this repeatedly with high profile breaches.

“Attackers are able to penetrate the datacenter, moving laterally and collecting information undetected, often for months on end (years in Yahoo’s case),” said Andrew Rubin, CEO & founder of Illumio, an “adaptive security” startup. “The ‘human middleware’ component of security operations is both untenable from a complexity point of view and near impossible to maintain from a personnel perspective.”

The problem is that security automation can’t remove this burden from security teams . . . yet.

“Next year, software algorithms and automation (the foundation for the AI that is sure to come) will play a greater role in all forms of cyber-security, from visualizing the activities of critical infrastructure we all rely on to enforcing policies that reduce the surface areas of attack available to hackers,” Rubin predicts.

Security is always an arms race. The bad guys are already using AI and are automating things like spam and phishing attacks. Security vendors will need to up the AI ante, using it to strengthen the weakest link in the security chain: us.

 3. Voice Assistants Begin to Mainstream Home Automation

Today, it’s no big deal to ask Siri to book a restaurant table or to request that Alexa call an Uber or order a pizza. In 2017, simple tasks like these will give way to more complex activities. As more people rely on AI-driven tools like Siri, Alexa, Cortana, and Google Now, automation will creep further and further into everyday life.

Mike Yurochko, CEO of Stringify, a startup developing software that “brings your physical and digital things together to create automated experiences,” notes just how pervasive smart devices have already become. “From a sheer volume perspective, lighting will continue to act as an entry point for first time smart device buyers.”

That’s right, lighting. Smart light bulbs that you can control with your smartphone and integrate with various smart-home assistants can be had for under $20. As the price continues to drop, soon pretty much every light bulb will be a smart one. Because, why not?

But it’s the all-in-one smart home products – such as voice assistants (Amazon Alexa products, Google Home, more coming) and multi-function smart cameras (Canary, Ring, Netgear) – that will really unlock value in the home automation space.

“Voice will be embedded into numerous entry points and will act as a consumer education point. This education will then open up opportunities for more comprehensive services that span beyond single point solutions,” Yurochko forecasts.

It may be hard to imagine how this will play out, but we can already handle many of what used to be called “daily chores” through apps and delivery services, so how much of a stretch will it be for a smart device to keep track of our preferences and patterns and start handling all of this for us?


In 2017, are we going to start being automated out of our own lives? Based some of these predictions, I’m beginning to wonder . . .