It’s been a big year for Artificial Intelligence, and its related terms like Cognitive Computing, Machine Intelligence, and Intelligent Machines, and for all its associated branches, from Machine Learning to Neural Networking and Natural Language Processing.
IBM, for example, has continued making strides with Watson in verticals such as healthcare and financial services. The vendor announced everything from advances in its ability to add rich image analytics with Deep Learning to the Watson Health platform, to a furthering of its partnership with Deloitte that will use the technology in solutions to more efficiently and immediately manage risk and regulatory compliance requirements.
At the end of last year, we also saw Tesla Motors CEO Elon Musk join with other industry figures to launch a non-profit AI research venture, OpenAI, with its stated goal being to help spread the technology to a broader base and direct it to having a positive impact on humanity.
Furthermore, just to pick one Google announcement in the area, the search giant said it’s using AI and Machine Learning in the Smart Reply capability in its Gmail Inbox mobile email client. Based on its recently open source TensorFlow Machine Learning platform, the feature reads messages and chooses an appropriate basic response for users.
This year’s off to a strong start too, with Facebook’s Mark Zuckerberg declaring his personal ambition to build an Artificial Intelligence system to help run his work and home life. He refers to it as “kind of like Jarvis in Iron Man.”
AI: Shake it Out
What else might we expect to see in the Artificial Intelligence (AI) world as 2016 ramps up – and even into the next few decades?
Take the AI open sourcing efforts, for starters. Pointing to the announcements of several open AI framework initiatives for different layers of the stack, from TensorFlow to Facebook’s Big Sur to OpenAI, technology futurist, entrepreneur, and angel investor Nova Spivack believes we’re just at the start of a trend. But, “while these initiatives are mostly open today, I would not be surprised if we start to see some (including new ones not yet announced) that have a hidden agenda to lock developers into competing ecosystems over time,” he says, noting that he discussed that idea here way back in 2013. It’s not a new strategy for major incumbents to release open source code biased towards their particular platform or ecosystem in order to gain strategic advantages, he explains, and some of these efforts have taken off, although he thinks developers today are wary of platform lock-in strategies, even when couched as open source projects. Though yet to be fleshed out:
“The OpenAI initiative is the only one I have seen so far that truly does not seem to have a trace of platform lock-in behind its agenda,” Spivack says. “If it is able to execute and develop technologies that developers actually need and adopt, it could result in the AI equivalent of Linux, and that would indeed be a very important development.”
Will there ultimately be one “AI OS,” or many? Spivack thinks it would be a lot easier to start with one that is open and essentially non-sectarian about which platform and ecosystem it lives in, in order to avoid the same many-to-many integration challenges that exist today. But “even if there are many, they will eventually be integrated because they will need to interface,” he says.
At Next IT, which has been developing intelligent interfaces for companies for almost 15 years and provides Intelligent Virtual Assistants to help businesses deliver online customer experiences, founder and CEO Fred Brown is thinking about the emergence of AI stacks that companies will put together based on their specific needs, “much like businesses build their CRM stacks today,” he says. To that end, it is critical for these various Artificial Intelligence systems to be able to work with each other. “We will have many companies that excel at specific functions and solution areas, working well together rather than a single dominant platform,” he thinks.
AI: Shake it Up
In 2016, AI is already in our lives more than we even realize, Brown says, such as in chat interactions and smarter self-service capabilities around customer service, where it found a home because it provided immediate ROI. We’re now seeing AI extending to the use of virtual assistants to augment overall customer engagement strategies, and he expects to see it deployed for workforce optimization too. Spivack agrees about intelligent virtual assistants being a big area of innovation, citing companies innovating in the area that include NextIT (which he advises), as well as Microsoft with Cortana, Apple with Siri, Google Now, Facebook M, and Viv.
Aggressive pursuit of Artificial Intelligence understandably has been undertaken in the healthcare sector, which is attempting to deal with massive scaling and disruptions, Brown notes; and, in financial services and travel, which have always been very much ahead of the curve in deploying AI. Various levels of Machine Intelligence also have had legs in everything from advertising with their targeting systems to IT with cyber-security initiatives. But, Brown says to look for it to start becoming more common across a number of verticals and industries where one wouldn’t typically expect AI to play a pivotal role.
That said, this isn’t the year that AI will be an absolute imperative for companies to pursue, as he sees it. Three to five years from now, though, they may wish they had at least started to think about it more seriously at this stage.
“Right now, the companies that are building their AI stacks are still ahead of the curve, and that is already paying off for them,” says Brown. “Those that are not yet investing in AI are already falling behind, and will be scrambling to catch up over the next year or two.”
And when AI does become an imperative, it will change everything. “This cannot be underestimated,” says Spivack. Think, he urges, of the race to the the self-driving car and how it “is going to have a truly massive social and economic impact.” Every function in every medium to large enterprise ultimately will be profoundly impacted by AI in the coming decades, Spivack says, in respect of job eliminations, job creations, and job productivity increases and in respect of business processes that will speed up and become more efficient and intelligent.
“Early adopter organizations that gain an early lead in adopting and integrating cognitive computing technologies into their business processes will have a first-mover advantage that could result in exponential advantages that late-adopter competitors will have difficulty catching up with,” Spivack says. “I think the real scare is mainly the risk of being late to the cognitive computing revolution that is going to sweep across all kinds of organizations for the next 10 decades.”
The good news is that in 2016, AI is still a real opportunity for companies, Brown believes. “There are still so many opportunities in how we think about AI, how it is integrated and what it can accomplish for the enterprise,” he says, adding that he would encourage companies not just to invest in keeping up, but to think about those areas where AI can be a real competitive advantage.
For pretty much every company, one such advantage likely lies in adopting Cognitive Computing technologies to deal with the massive growth of unstructured and streaming data.
“Humans simply cannot cope with the volume, velocity, and variety of data that is coming at them today. And we are only in the beginning of this growth curve,” says Spivack. “We are going to need AI to process and make sense of all this data in motion — not just in real-time but even after the fact — because it simply is too vast for the human brain to comprehend.”
That perspective has informed Spivack’s co-founding of Bottlenose – where he also is CEO – which provides a truly continuous Cognitive Computing stack designed to make sense of unstructured data at scale.
AI: Shake and Serve…But Carefully
As Artificial Intelligence moves forward in 2016, integrating further into business practices and people’s everyday lives, Brown suggests that developers, designers, business execs, and users pay close attention to “how we want to interact with AI….The decisions we make now about design and interaction will determine a lot of how the technology will develop, how we will use it in the near-term, and even what functions we prioritize.” As we develop and deploy more and more powerful AI, keep in mind that the work is as much about creating technology that can advance human and machine interaction as it is about machine-to-machine interactions, he says.
Spivack also is concerned that the use of AI in autonomous weapons systems such as drones, military robots, and intelligent missiles may proceed without the safeguard of having humans in the loop when lives are at stake. “Unfortunately, we may see high-frequency warfare (as in high-speed, AI-powered) emerge – powered by automated military systems,” he says. “Humans need to be the failsafe in these systems.” After all, there’s a place for doubt when it comes to assessing sensor data and a place for relying on gut instinct that could keep a bad situation from getting worse. Humans can bring both to the table, but “AI systems don’t have gut instincts, intuition, or respect for life and they are not good at doing illogical things,” Spivack says.
These are serious risks, but by and large Spivack thinks that the change to our lives that AI will bring is going to be relatively ubiquitous over time as tasks and processes that used to be labor-intensive become semi- or fully-automated and productive. The change, he says, is going to be primarily generational in timescale.
What’s that mean? Think of it this way: The first babies of 2016 whose photos were splashed across web sites and newspapers in their respective areas of the world last week are going to grow up with Artificial Intelligence. It will be part of their daily life, they will rely on it, and they won’t find it at all surprising or unusual, Spivack says. Indeed, they’ll be hard-pressed to understand how anyone could have lived without it. For us adults, it’s a different story.
“[We] are going to experience the most technological shock,” Spivack predicts. “When we are near the end of our lives the world will be drastically different than the world we were born in. We will certainly see the radical changes we have witnessed in our lifetimes as amazing. But to our kids, they will be just the way things are.”