Artificial Intelligence and The Future of Tolerance

As Trump’s victory in the US, Brexit in UK and the rise of far right in Europe recently demonstrated, tolerance is globally on decline. There are multiple reasons on how we got here, and I think use of artificial intelligence is one of them. I also think it can potentially save humanity.

Artificial what?

Historically, computers have been good at tasks we humans are not particularly good at – such as calculating big numbers – whereas they have failed notoriously in tasks that come natural to us, like recognizing cats in photos.

In the recent years, however, storing and processing huge amounts of data became not only possible but also feasible. The marriage of computer science and statistics, fueled with more data and processing power than ever gave birth to the AI revolution: By borrowing ideas from how our brains work, scientists were finally able to use statistical models on huge datasets to build neural networks and train computers to play games, recognize patterns and learn.

Artificial Intelligence is important because it allows us to build better cities, lead healthier lives and improve our financial systems. It’s also important because it will replace our jobs, summon the demons and bring an end to mankind.

What is AI doing to us?

The reasons why we have (so far) been better than computers at pattern recognition and deduction is purely evolutionary: if our ancestors had to sit down and do a rational analysis every time they heard a growl behind a bush, they would be eaten alive by tigers. Our ability to detect patterns helped us survive in the savanna.

However, in the past few centuries the world has changed must faster than our primal brains could adapt. Our tendency to take mental shortcuts which helped us survive in the the ancient world morphed into logical fallacies in our increasingly complex modern world. It started hindering our ability to make decisions on complicated topics such as how to build our financial, political and economic systems, deluding us to mistake correlation for causation or to see patterns where none exist.

One of these fallacies, confirmation bias, would be familiar to anyone who has ever been exposed to a mindless presentation in a corporate meeting: it makes people search for information that reinforce our preexisting beliefs, while actively discarding opposing views.

This is where AI comes in: scientists, designers and developers have figured out how to build addictive systems that feed our bias for confirmation. Why? Because that’s where the money is.

Enter Skynet

We may not even realize when algorithms help us decide what to watch, get a good deal or avoid traffic jams. However, the implications are much bigger when it starts to shape our perception of the world.

AI has had huge impact in social media and journalism. The business model for social networks is advertising. The main success metric in social media is “engagement”, a representation of its users’ fragments of attention measured by clicks, likes, comments and shares.

By constantly collecting data and tweaking their algorithms, social media companies did an exceptionally good job at keeping us on their apps and platforms, feeding us more of what we respond to and making a lot of money by selling ads in the process. This is why social media companies actually has very little interest exposing us to different points of views.

It would still be fine if we were just using social media for looking at funny cat videos or keeping up with old friends. However it has become our main source of receiving news, algorithmically deciding what we should see.

Add psychographic profiling and thousands of data points on potential voters, personalized messages designed to trigger an emotional response delivered through social media, and you end up with a tool so powerful that can even effect the outcome of an election and change the world.

The economic models surrounding the use of AI – especially in social media – feeds into our echo-chambers, further fragmenting our societies into self-serving bubbles.

And when we move from smartphones to virtual reality and become engrossed in it, it will become even easier for us to lose our reference points for facts and reality.

So these alternate realities with no exposure to different points of views breed the ideal conditions for skillful manipulators to drive a wedge between the cracks within the societies, pushing them further apart.

What’s next?

For the first time in history, we have the right technology and the ingredients to build truly connected societies, bridging the gap between different world views, replacing separatism and tribalism with open-mindedness and tolerance.

Silicon Valley frequently talks about making the world a better place. I think the the right way to actually do it has less to do with dominating markets and more on figuring out how to build sustainable business models that bring societies together.

I think there are three main areas where we start using technology and AI to build a more tolerant future:

Education is the fundamental building block for understanding our world and opening our minds, hence should be our starting point. The problem is that most of the time it is dreadful. I think AI has a role to play in shaping the future of classrooms and make it fun again. Alt School is one of those initiatives that tries to build custom curriculums through smart use of tech and data. Carnegie Learning offers personalized math instruction for grades 6-12. As the use of technology and AI shapes the future of classrooms, I think we will see a transformation from rote to personalized learning and as a result, more engaged and more open-minded students armed with the joy of learning.

Travel exposes us to different cultures and expands our horizons. With the decline in cost of air travel and the growth of online travel agents, we can think of ways to use technology to encourage and inspire people to travel more. Companies like WayBlazer use AI to discover traveler intent using hundreds of data points and connects it with the right accommodation. The same technology can be used to gently nudge people with travel intent into exploring new experiences. Airbnb recently announced “experiences”, allowing travelers to book local activities and get immersed in local communities. As more users and brands turn to social media for travel inspiration, I’m excited about travel industry’s ability to drive meaningful change in opening our minds.

Journalism helps us uncover facts and keeps organisations honest. However the advertising business model for news outlets in the age of social media is in direct conflict with it’s original goals of journalistic integrity, which is why we can’t stand news anymore. News outlets should figure out how to build a sustainable business model that support their original goals (hint: subscription) that does not rely on shoving ads in our faces, so we can maintain our reference points for facts and reality. Although it’s easier said than done.

I don’t know how future will unfold, yet I do feel we have both the tools to shape it as well as the responsibility to be mindful of the consequences while doing it. This means considering the social impact of the businesses that we build as entrepreneurs, or the nature of businesses we choose to support as consumers.

How do you think technology and AI can contribute to building more open-minded and tolerant societies? I’d love to hear your thoughts.

Tinder and The Sharing Economy

Platforms like Uber and Airbnb have been tremendously successful, precisely because they are extremely effective at connecting supply and demand. So good that they make us question the lifestyles we’ve taken for granted: Do we really need to own a car when we can Uber anywhere or buy a house where we can Airbnb our way around the world?

Tinder is suprisingly similar to Uber and Airbnb in the sharing economy: it’s a novel approach to catering to our fundamental need for human connection, physical or emotional. It invites us to question the expectation that relationships should be long term and monogamous.

The idea of romantic marriage is relatively new. We now turn to one person, asking them to give us what an entire village used to provide not so long ago. No wonder we are so bad at it.

Do we really need to be in a committed relationship if we can have intellectual stimulation with one partner, mind-blowing sex with another, and emotional fulfillment with a third?

Are we returning to communal societies weaved together by smartphones and fuelled by smart platforms, where the tribe takes care our needs, whether transportation, accommodation, sex or companionship?

Leaving the judgement on how to live our lives “the right way” aside, is this transformation is happening so fast that we are failing to notice?

Personalize or Die

Attention is a scarce resource. Everyday, brands are fighting a decreasing portion of the attention pie, with significantly increasing costs. Although this one-upmanship is great for ad networks, it’s destructive for your bottom line.

Repeating your message over and over — hoping that if you do it for long enough, somebody will eventually pay attention — is useless and expensive.

Smart brands realize that collecting, analyzing and making use of customer data is great for building brand equity and loyalty. Breadcrumbs of valuable customer interaction at every touch point allows these brands to provide more relevant and timely product and services.

For example, Amazon has consistently improved their product recommendations over the past decade. Netflix granted $1M in 2009 to a team that were able to improve the accuracy of their recommendations by 10.06%. Target is able to identify pregnant women in their third trimester.

One of the most exciting updates is from Facebook: They recently made it much easier for themselves to analyze the data of their 800M+ users by introducing verbs in status updates with Timeline feature. Research suggests that when customers go through major life events, their habits become flexible. And naturally, Facebook is positioning itself to sell this window of opportunity at a premium to its advertisers.

Arguably, one of the main reasons why smaller companies do not invest time and money into providing a fine-tuned level of personalization is the complexity of the task, and the perceived (lack of) value it will bring. However, it’s never too early to start.

Tools like KISSMetrics for customer analysis, Custora for CLV calculation have made it affordable and easier for you start understanding your customers. Tech startups such as Platfora promise a brighter future for marketers. For a technical introduction to your home-baked solution, Programming Collective Intelligence is a good start to understand recommendation algorithms.

Personalized marketing is the future, and it’s here. Start looking for patterns. Figure out what makes your customers tick. Use your findings to tweak and improve your products and services. Repeat.