— An analysis of our debut event on AI, the technology that walks the fine line between a world of wonder and one of woe. From cybersecurity to healthcare, discover the nature the most polarising technology of our generation.
On February 10, 1996 the first chess game between a human champion and a computer took place. Garry Kasparov, the international champion at the time, kept the machines at bay, beating IBM’s Deep Blue five games to one. A year later, Kasparov lost a rematch to Deep Blue. Not only was it his first match loss ever, but it was also the first time a computer won a game against a human in a tournament.
Years later, Kasparov said that even if he overwhelmingly won against Deep Blue in 1996, he understood that something was changing. If a computer can win against a world champion even once, then soon computers would have the ability to win every time. In this sense, 1996 was a huge milestone: Garry Kasparov was the last human champion ever and since then moment computers have always won.
Artificial Intelligence is the technology that redefines the boundaries of human potential and has the capability to transform the ways in which we interact with each other and the environment.
There are people who think that super intelligent humanoids could be humankind's last invention, whilst others think that AI will unlock a ever-lasting wave of new opportunities. From enthusiastic tech gurus to sceptic scientists and prudent experts, everyone seems to have their own point of view on this transformative technology. It was this which our debut event, maize.live - The Rise of Artificial Intelligence was dedicated to.
Throughout two days of in-depth analysis, inspiring debates and thought-provoking speeches, we delved deep into the implications of Artificial Intelligence and the future of humankind.
Exploring the dark and the bright sides of this technology, we learned that our collective future is exclusively in our hands. The way we shape our tomorrow depends on how we face the enormous change that AI is bringing into our lives today.
So what were the most powerful insights on Artificial Intelligence gained?
Robert C. Wolcott, Co-Founder & Chairman, The World Innovation Network (TWIN) Clinical Professor of Innovation & Entrepreneurship at Kellogg School of Management Managing Partner at Clareo, moderator of our event maize.live - The Rise of Artificial Intelligence
“There’s no magic, there’s nothing special: it’s basically simple math” - as Pascal Weinberger, Head of Rapid Development & AI at Telefonica Alpha Innovation, suggested, in short, let’s stay grounded. Even if AI might worry us, it is unlikely that this technology will suddenly revolt and take over the world. Despite the fact that artificial systems are growing quickly, they are not human and are based on mathematical structures. Human intelligence is extremely complex, multifaceted and involves many areas such as common sense, reasoning, language, planning, analogy. Today, the only field in which AI really can make the difference is deep learning. Machines can't do anything on their own and their decisions and actions are still trapped within the limits of what we deem necessary.
Rather than seeing it as a threat, we should view AI as a field of limitless possibilities. Artificial systems learn from the data we give to them, if we want to create powerful machine learning algorithms, we have to choose correct, positive and bias-free data, keeping in mind that the inputs coming from the surrounding environment are actually influencing the way this technology makes decisions. As machines become more and more sophisticated, the real challenge for us is to define how we program and employ these systems. If we always keep humans in the loop, AI will neither surpass nor defeat humankind.
“The best way to predict the future, is to create it!” - Abraham Lincoln
Today, crime is being lead by highly intelligent and extremely smart people. According to Menny Barzilay, CEO and Cyber Security Strategist at FortyTwo, crime itself is getting more innovative and creative every day .
In the world of crime, we can recognize an inherent asymmetry: cyberspace intrinsically favours hackers and hinders security people’s work. As with terrorism, if you are a hacker you only have to succeed once, if you are the security agent you have to succeed every time. Also, security is very costly, hacking is very cheap. Artificial Intelligence can change the rules of the game and disrupt this pattern, contributing to cybersecurity in fields such as anti-spam and phishing, automation, malware detection and many more.
Innovation is a two sided coin and every new technology creates new problems.
Some years ago, the Defense Advanced Research Projects Agency (DARPA), the agency in charge of developing defense products, systems and technology for the USA, organized a competition wherein participants tried to hack very complex defense systems. For the first time in history, the participants weren’t human: machines competed against other machines with AI replacing human hackers.
In this unsettling scenario, where AI systems acted simultaneously at the same time, both as the threatening actor and the “good guy”, a question arose: who is more likely to win? Artificial systems would most likely maintain the same asymmetrical pattern of current security systems and it will, once again, prove to be harder to guarantee security and easier to commit crimes.
Today, algorithms can learn to ID pixelated faces and even evaluate someone's personality and psychological traits. From AI-assisted fake porn to real-time face reenactment, we are able to make these machines truly infallible. AI systems can mystify our perception of reality and we are at risk of losing the ability to understand what is real. Trust is our only weapon against this tendency and without it we can’t move forward. Trust is something we have the power to create and it’s imperative to work together to make sure that AI does not divide us.
“AI will create amazing, amazing opportunities for humankind, but also amazing, amazing threats.” - Menny Barzilay
AI today is very powerful but remains, ultimately, a black box: you have an input, something happening in between, and then the output. As with us, something occurs within the AI brain/engine that we are not privy too: we don’t really know what is going on or how decisions are being made.
However this is going to change, in Europe at least. The new regulation which came into effect last week, GDPR, stipulates in art. 13 that it requires anyone who is using AI to know why and where certain decisions have been taken by the machines. Businesses are finally being called to find a chain of accountability.
Starting from this urgency, Professor Mischa Dohler told us how in King’s College London, a new concept of explainable AI planning (xAIP) has been developed. This concept works on a very deterministic decision tree where causality is explainable and decisions made by the machines are re-computable. The most interesting aspect is that these explainable AI systems can coexist with our decision-making process. Humans can tell the robot to change a decision taken and the computer can start to act from that point the decision has been taken. This results in a much more accountable, open and traceable type of AI methodology.
“AI will automate jobs, and humanize work.” - Mischa Dohler
In spite of the hype, Artificial Intelligence it’s not about technology taking over, but rather about extending our limits. Exponential technology can work as an extension of ourselves and capabilities that help us to solve complex issues. As a good example of this potentiality, Bart de Witte, Chair Faculty of Digital Health del futur/io Institute, shared with us the powerful revolution occurring in healthcare that we can trigger today simply with a tap on the smartphone.
Overperforming algorithms, data sensors and AI together are creating huge possibilities for early diagnosis, but it is the combination of human intuition and judgement, together with the precision and consistency of these machines that will supersede the performances of both. The future of healthcare is not humans versus algorithms, but humans and algorithms against disease.
The extraordinary power of these technologies also stems from the potential to be open and accessible to everyone. 70% of the world’s population has no access to healthcare – AI could provide key solutions to this huge problem, saving millions of lives. Healthcare for all is a desirable future and humans play a central role in shaping it.
While algorithms can automate many aspects of our worklife, the nature of artificial systems itself implies that humans will always be central. Scott David, Head of Information Interaction at the World Economic Forum, explained that algorithms should be viewed as cognitive tools capable of augmenting human skills and redesigning organizations.
We should work not only using AI to optimize aspects such as customer experience or supply chain, but also on scaling our own cognitive capabilities and augmenting people and jobs. To make the most out of AI technology, a design process is essential in driving innovation across organizations and defining our relationship with these machines.
AI should become an extension of the individual. While computers are better at making predictions and calculations, people are better at rethinking experiences and redesigning processes, modelling our world in order to convert human thoughts into a human-centered data structure.
“Technology is giving us the capacity to become superhuman.” - Bart de Witte
# CUSTOMER CENTRICITY
In almost every industry, AI is entering corporate DNA – revolutionizing business models and dynamics. With Lars Schwabe, Associate Director of Lufthansa Industry Solutions, we investigated how Lufthansa is focusing its attention on the design of customer-centric experiences in the age of AI. When it comes to improving the customer experience, conversational interfaces are gaining momentum due to their ability to simulate conversations with humans and offer instant and effective digital solutions. A good example is Lufthansa’s bot, Mildred, a tech savvy assistant that you can contact easily via the Facebook Messenger App.
An increasing number of customers prefer to deal with chatbots in certain situations to obtain immediate information and fast answers to their needs and queries. To engage with customers in a way that traditionally only a person would do, chatbots rely on Natural Language Processing (NLP). Today, Lufthansa is working intensively on these aspects, developing business-relevant customized NLP solutions.
We now see humans and AI merging their strengths to achieve the best result for all involved: while chatbots can handle a considerable amount of repetitive tasks and analyse complex situations, humans can focus on higher value services, therefore providing an overall significant and effective customer experience.
Another interesting example of how artificial systems can assist humans in high pressure and complex situations was given by Vittorio Di Tomaso, H-FARM’s Artificial Intelligence Director. By using machine learning to filter information and detect patterns, we can predict natural disasters and improve the way we respond to them.
Extreme weather events are expected to become increasingly frequent and longer in the future. Extracting information from the large quantities of available data from different sources, including social media and crowdsourcing, we can extrapolate and provide fast and effective information (as well as solutions regarding prevention) in order to quickly react to natural disasters.
This is how the European project I-REACT (Improving Resilience to Emergencies through Advanced Cyber Technologies) is taking advantage of predictive AI. Thanks to mobile apps, social media analysis tools, drones, wearables to improve positioning, and augmented reality glasses to facilitate reporting and information visualisation by first responders. All of which allow actors such as organizations, policymakers and stakeholders to improve disaster prevention and response.
“Improving Resilience to emergencies through Advanced Cyber Technologies.” - Vittorio Di Tomaso
# BIG DATA
Among the financial services, insurance is one of the key sectors that is exploring the possibilities of AI. Traditionally, a customer only thinks of insurance when approaching a purchase, a significant life change or when facing an unfortunate life event. Today, insurance companies are trying to create a different perception of their services, becoming active participants in the customer's daily life.
As Reza Khorshidi, Chief Scientist of AIG, told us, the insurance sector is trying to increase its proximity to customers and reduce the traditional gap between brokers and clients, taking advantage from the spread of the ecosystem model. This trend explains how engaging with a customer on a need other than insurance (e.g. travel, healthcare, entertainment), by gathering large amounts of data and understanding in-depth customers’ needs, insurance companies will be able to improve their speed and accuracy in creating more personalized and direct-to-customer offerings.
Data is also central to new fintech realities, as Roberto Mancone, Chief Operating Officer of we.trade Innovation DAC, pointed out. Banks are already exploiting the power of data and advanced analytics to identify, segmentate and acquire customers, keep track of the customer journey, improve credit risk decision making, develop and enhance SME behavioral rating models (using social media data) as well as early warning default detection.
“Software is Eating the World, But AI is Going to Eat Software.” - Jensen Huang, CEO, Nvidia 2017
A true, meaningful journey through Artificial Intelligence, its wonders and threats, leads to an obligatory question: how will this technology effect us as human beings? Our last speaker guided the audience in reflecting and identifying the most potent considerations and questions for the future of human culture, behaviour and even the ways in whic each one of us thinks as individuals.
With a Neuroscience Ph.D. and a past experience as a science journalist and author, Lone Frank pointed out the fact that technological innovation is deeply connected to two great “mental revolutions”, comparable to what Copernico and Darwin’s theories did to transform our view of the world forever.
On one side there is neuroscience: for centuries we have been thinking that there was something immaterial and solely human that defined our being. Now we know that is the “state of our brain” that define us, a unique organ that can be empowered through the use of new technological implants and substances. On the other side there is genetics, and the ongoing popularization of genome sequencing.
But while we are still discovering more and more about our own mechanics, without concepts such as “soul” and “inner personality” how can we define “who we are?”.
AI is the latest ingredient of this revolutionary recipe. Machine learning and data analytics are unveiling the fact that human behaviour is predictable and that we are almost biological machines, with inner rules and patterns. So, if a machine could learn these behavioural norms, what remains that is human? What is our role in society if not only our jobs can be completed by a piece of algorithm, but the way in which we think itself?
Cogito, ergo sum?
“Our behavior is predictable, the more we use AI the more this will become clear. And in the longer term this fact will shape human beings, as individuals and our society” - Lone Frank, Neuroscience Ph.D., Science journalist at Weekendavisen
In case you missed it, you can check out all of the action here.