“Humans and machines could merge so completely that humans might not survive if they went offline,” said Yuvul Harari, author of several bestselling books that are beloved in Silicon Valley, including “21 Lessons for the 21st Century”.
Welcome to the fourth Industrial Revolution – the era of Artificial Intelligence, where most of us are not even aware that our lives are so tightly interwoven with algorithms. Algorithms constantly scan our biometric data and emotions, dictating what we think about, talk about and consume. Without realising, we are selling our souls to the devil that is Big Tech companies and the government who could manipulate us and make decisions on our behalf, says Harari. However, we can’t honestly say that we are being deceived since the Facebook-Cambridge Analytica data scandal and “The dilemma of social networks” documentary, among others, have already warned us of data theft, political interference and blackmail that run behind the thick veil of the virtual world.
The truth is that society as we know it is evolving at such an accelerated pace that we are unable to determine if the technological disruption we are experiencing will bring more benefit or harm. So, is Artificial Intelligence closer to being a “Vision” or an “Ultron”? In other words, is AI a hero or a villain? According to Dr. Mark Esposito, author of AI Republic: Building Links Between Humans and Intelligent Automation, it seems that AI is neither one nor the other.
Do we really understand what Artificial Intelligence is? Dr. Mark Esposito breaks it down for us.
Artificial Intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. Though of course, it can’t do what the brain does, Mark explains. So while the AI can work like ten brains to quickly calculate statistics for when it will rain, for example, it won’t be able to do something different like sing in the rain unless it’s specifically programmed to do that.
While we envision hover cars and friendly companion androids, Mark says, “it’s not that we’re going into a Futurama state, but we’ll have a much broader integration of AI solutions into our day-to-day lives”. On a macro level, there is AI integration for specific businesses. On a micro level, we will experience more AI in our mobile applications, health watches, or virtual reality headsets.
There are already documented cases of how social networks feed us the information they believe we are more compatible with and therefore more likely to consume. This feeds into an echo chamber which reaffirms an individual’s beliefs, further polarising their opinion, which was how the victories of the Trump and Brexit campaigns were conceived. On a lighter note, there is also a certain degree of manipulation in Netflix and Amazon’s movie and product recommendations.
The main limitation of Artificial Intelligence is that its learning is only quantitative. The biggest challenge now is for AI to move beyond the quantitative, towards understanding context and social science disciplines using qualitative methods to enhance its value. This is something humans can easily do that machines cannot, explains Mark. That is why translation application programming interfaces struggle deeply with certain expressions and words out of context, which only improves as we ‘suggest changes’ in certain situations.
In his book, Dr Esposito mentions that machines are very fast with numerical data, but the visual field is more difficult for them to learn. To understand this, let’s use our phone’s facial recognition function as an example. To unlock, we have to first configure the facial recognition by rotating our faces in different directions, whilst our phones capture the images. Things were working well until our phones were unable to recognise us with our masks on, rendering the function impractical and almost useless if its only job was to recognise our face. This is something which we humans still have an advantage over.
The truth is that Mark affirms that the opportunities of Artificial intelligence are endless so that he can be both a hero and a villain. On the one hand, AI can help us gain big wins in finding medical solutions, developing systems that allow us to be more connected, and models that help governments boost vulnerable economies. In short, making the difficult easy. However, AI’s potential as a villain will be determined if the objective sought by the programmer or owner of the technology is to extract value from consumers, increase inequity and disparity, and weaken or manipulate people.
This blog post was translated from ejecutivos.es, featuring Mark Esposito.
1 St Katharine’s Way,
E1W 1UN, London, United Kingdom
6 Battery Rd, #03-68,
The Work Project @ Six Battery Road
Studio 1006, Dreamplex Thai Ha, 10th floor 174 Thai Ha Str, Dong Da Dist
Hanoi 100000, Vietnam