This is how AI can help you make sense of the world

(ITU, 2017)

This article is brought to you thanks to the strategic cooperation of The European Sting with the World Economic Forum.

Author: Sean GourleyFounder and CEO, Primer

Our world seems to be increasingly volatile. Geopolitical events occur that we can’t believe, global protest movements seemingly emerge from nothing, and new technology comes online every week that seeks to upend the way we’ve done something for decades. Events unfold at a pace and complexity that we as humans can’t comprehend.

However, the world is actually not as chaotic as it can sometimes feel.

There are patterns around us, but we can’t see them because we’re biologically constrained by a limit to the volume, speed and complexity of information that we can understand. Scientists find that when we remove these human limitations and employ machine intelligence to process data at a super-human scale, then structure and patterns emerge. Chaos gives way to clarity.

My PhD research focused on the mathematical patterns underlying war – patterns that predict the timing, size and geographic distribution of attacks. Similar patterns emerge in other complex systems if we train machines to process, read and understand enormous volumes of information.

We’ve been exploring this with our commercial work at Primer. Banks, intelligence agencies and global retailers are using our software to make sense of the world, because our cognitive abilities can’t keep up with the information produced today.

Machines will reduce the cost of curiosity

The most interesting thing about Artificial Intelligence is not its ability to mimic human intelligence, but its ability to see the world differently than us. Machines are a different kind of intelligence, and as such, they see and understand things that we cannot. The key is for these machines to teach us what they understand and allow machines to share their insights with us.

For example, what if you wanted to quickly grasp the most important events in the women’s rights movement of the last year. While human precision is high, our recall is not. You might be able to recite a handful of events or people associated with the movement, but you would likely miss many details of the key events, stats or people that shape the entire narrative. You almost certainly wouldn’t be able to recall events that took place worldwide. And given your own unique bias – the news streams you follow, your geography, political leanings and age – you couldn’t construct a complete picture of what’s unfolding.

The same is true for expert analysts that cover these kinds of events. If you were to set out to do research, it could take you weeks to sort through all of the data to accurately identify the key moments, actors, numbers and how they’re all connected.

Machines can do all of this in minutes. Our software was able to mine all of the 33,000 English-language documents in our data set associated with the women’s rights movement. Using machine-learning algorithms, Primer identified, timestamped, described and geo-located all of the events happening around the globe.

This series of events can then be compared to every other political protest in the data set to uncover patterns in the way that events unfolded.

Primer’s visualization of key women’s movement-related events around the globe extracted with AI from over 33,000 English-language news and blogs.

Now, imagine you’re an analyst, digging into something more complex, like Chinese language briefings from the past decade on political events in Shanghai, or private data on clinical trials of every cancer drug ever tested from the last five years. The human requirement is far too costly and inefficient for most organizations. And yet we’re sitting on troves of useful but untapped data that may reveal answers.

Now imagine that you were able run this analysis for all information, all the time. You would be able to construct a continuously updating, human-readable representation of the current state of the world.

We’re still in the very early days, and it’s already possible for a machine to read all of this data and surface the most meaningful events, places and people, so that you can draw connections and conclusions about the world.

Today, it’s possible for AI to create the first draft of a Wikipedia page with greater recall than a typical human editor and comparable precision. As AI and natural language understanding improves, the outputs and benefits will be far more meaningful.

Imagine if you had a team of the world’s best analysts at your disposal, all day, every day. What questions would you ask? What would you try to understand, that seems impossible to understand today? What seemingly random connections would you explore, and where would your curiosity take you?

If we all had access to augmented intelligence, we would have a world where state officials and the intelligence community will use accurate data, executives could make critical decisions that could impact global sales, elections wouldn’t be influenced as easily by propaganda, and your average person could make more informed decisions about politics or purchases.

This is where the world is headed. AI is on the cusp of materially changing our own intelligence and decision-making ability – and I believe for the better. Just as we saw the replacement of mechanical, human-muscle labor with machines during the Industrial Revolution, we’ll see a similar disruption in the modern workforce. Artificial Intelligence will bring economic and societal disruption that could be more significant than what we saw in the Industrial Revolution. There’s also no question it will be the next offset for how wars are fought and won.

This is the natural cycle of evolution – periods of rapid technological advancement followed by plateaus and more rapid advancements. But without this mechanization of cognitive process, we simply won’t be able to navigate this world. The net gains society will gain from artificial intelligence far outweigh the dangers.

The challenge for any organization in the next decade will be integrating machine intelligence. The better you are at assembling teams of humans with machines, the more successful you’ll be. If you make a bet with machines-only or humans-only, you’ll miss out.

AI will surface truth

The benefits of a world with augmented intelligence are undeniable. It’s paramount that we find patterns and truth in the world that humans alone can’t surface. In the future, those with knowledge will have power. And knowledge will lead to truth.

AI, machine learning and natural language processing will help us make sense of the seeming chaos around us. The more sophisticated the tool, the more sophisticated our understanding of the world will be.