What do you get when you cross a colossal and complex amount of data with computing technology so advanced that it can outperform humans in almost everything? Depends on who you ask.
“You can have data without information, but you cannot have information without data,” says computer programmer and science fiction writer Daniel Keys Moran. And that’s a problem throughout the world right now: the so-called “big data” monolith. Modern technologies have succeeded in gathering so much data about so many topics that no human can efficiently analyze and interpret it all.
As a result, people are turning to artificial intelligence (AI) to address this predicament. Because computers can read and analyze data exponentially faster than humans, the AI approach is tailor made to sift through terabytes of data and deliver the valuable information upon which sound decisions can be made.
So why the alarm bells surrounding the fusion of big data and artificial intelligence? AI has been identified as “our biggest existential threat” by one of the greatest minds on the planet: Elon Musk. And Stephen Hawking, one of the world’s most admired scientists, told the BBC, “The development of full artificial intelligence could spell the end of the human race.”
Exactly What are
We Talking About?
Before delving into these apocalyptic issues, it is important to define these ideas more clearly. “Big data” is the term used to describe the collection of raw facts and statistics that have been assembled in any given subject area. Not only is big data characterized by its size, but it also refers to data that is exceedingly complex and/or available from an immense number of sources.
In this context, “artificial intelligence” is the tool that is used to tame big data. AI systems are designed not only to sift through and perform simple analytics on large caches of data, but also to find patterns, reach relevant conclusions, and apply these results toward advancing specific goals.
The aspect of artificial intelligence that has led to the most distressing rhetoric is known as “machine learning.” A subset of AI, machine learning is the process by which computer-powered “machines” are able to evolve as they are exposed to new data, instead of simply performing the same preprogrammed instructions repeatedly.
It is this evolutionary property of AI through machine learning that has some experts worried. For instance, an AI-powered computer program recently played 60 games against the world’s top players of go, the most complex board game in existence—and posted a record of 60–0. Another AI program stunned skilled poker players in January by crushing them in Texas Hold ’Em, in large part because the program learned how to bluff. If AI can boast these achievements now, doomsayers claim, then what will it be able to do in the next decade or two?
Are there “things’ on the internet?
One thing is certain: AI is already impacting the day-to-day lives of businesses, laboratories, governments, and consumers. Computer users can see the fruits of AI almost daily in the various “ad choices” that pop up on websites, the tailored marketing messages they see on Facebook and Twitter, and the “products you might also like” that are displayed during e-commerce transactions. These user-specific links are offered based on an individual’s web activity and online preferences as determined by AI programs that monitor and analyze these countless data points.
And it is not just consumer data being targeted by AI. The Internet of Things (IoT) refers to the group of traditional machines, devices, and other entities that are now being equipped with electronics, sensors, and computer connectivity. These so-called “smart products”—which encompass everything from light bulbs to warehouses—can automatically collect and process data and then take a predetermined action. Many of them are found in “smart homes,” which can turn off lights in empty rooms, adjust the interior climate, and send video to a homeowner’s smartphone whenever someone approaches the front door. In addition, municipalities are embracing the IoT concept for applications ranging from identifying open parking spaces in downtown areas and automatically illuminating streetlights at dusk to real-time water quality monitoring and automated electricity meter readings.
All companies great and small
So how will the big data–AI merger change the world of business and commerce? One direct consequence will be the elevation in importance of the chief data officer (CDO). Many companies already employ CDOs who largely focus on compliance issues and regulatory documentation. But going forward, the role of the CDO will morph into a proactive function that will involve gathering and interpreting data for strategic purposes, as well as using data to drive innovation, increase efficiencies, and improve customer experiences.
Much like the established software-as-a-service (SaaS) industry, data may emerge as a service purchased from a community of cloud-based providers. For many companies, it is more cumbersome to allocate resources toward collecting and analyzing accessible but complex data than simply to pay another company to do it for them.
While larger companies will certainly facilitate the union of big data and artificial intelligence, many smaller, highly specialized firms will also play a huge role. To be sure, the giants of Silicon Valley have been experimenting with AI in search engines, electronic assistants, and many other applications for quite some time. On the other hand, there are dozens of startup companies in the healthcare sector who are trying to leverage AI and big data to improve diagnostic imaging, detect early warning signs of patient deterioration, and design personalized health management systems. And energy companies are now relying on AI programs to sift through massive quantities of geoscience data to locate the most profitable extraction sites and avoid the financial consequences of drilling at a dry site.
The end of humanity
as we know it?
Despite the myriad of advantages that could spring from a seamless unification of AI and big data, there are those who worry about the rapid advances in machine learning and how that might affect AI programs in the future. Inventor and futurist Ray Kurzweil predicts that “artificial intelligence will reach human levels around 2029,” while Shane Legg, the cofounder of British AI firm DeepMind, is on record as saying, “I think human extinction will probably occur, and technology will likely play a part in this.”
To be sure, the possibilities for misuse of these technologies are endless. Companies or governments may attempt to exploit the data they obtain by engaging in “persuasive computing” to shape the flow of information to boost profits or achieve political goals. For example, a Chinese search engine is teaming up with the military to analyze online user data, with the end goal of assigning a so-called “Citizen Score” to every person, which would impact the jobs they can get, the loans they may qualify for, or the travel visas they can obtain.
To combat these potential calamitous outcomes, Scientific American has proposed a set of fundamental digital principles. Some of them focus on the information itself, including decentralization of information systems, reducing information pollution or distortion, and putting information filters in the hands of users. These precepts also call for improved transparency and interoperability, an emphasis on collective intelligence, and the promotion of responsible digital behavior of entities who rely on artificial intelligence and/or big data.
The major players in artificial intelligence are also starting to recognize the undesirable potential of unbridled AI development. In September of 2016, the largest tech firms in the United States founded the Partnership on Artificial Intelligence, a group designed to tackle AI-related issues with the goal of avoiding dire consequences. Even so, these companies and others will not be hindering the merger of big data and AI anytime soon because of a few ominous warnings. Indeed, Mark Zuckerberg urged attendees at a 2016 conference to “choose hope over fear.”
