This week, we analyze how AI (Artificial Intelligence), a topic so widely covered by media, is influencing our everyday life.
AI is being implemented in your mobile device, on your favorite social media streams, we even reported about AI implementation in Virtual Anchors. There is no doubt a high demand, but one thing is certain. It is the future!
According to a new report published by Allied Market Research, titled, “Artificial Intelligence Market by Technology and Industry Verticals: Global Opportunity Analysis and Industry Forecast, 2018-2025,” the artificial intelligence market accounted for $4,065.0 million in 2016, and is expected to reach $169,411.8 million by 2025, growing at a CAGR of 55.6% from 2018 to 2025. You can read more here...
When it comes to latest technology, you can of course ask any tech guru. But when it comes to being among the first to implement new technology, you can ask Angelo Dalli who you can refer to as an Entrepreneur, Investor and Technology Expert. (make sure you connect with him LinkedIn or Twitter).
Angelo Dalli has been creating innovative technology companies since 1995, in the fields of search engine technology, intelligent transport systems, online gaming, AI and big data. As an entrepreneur, Angelo has had a number of successful exits including M&As and listings on European and US stock exchanges, is a member of EBAN, and has also received the Top IT Entrepreneur award in 2014 in Malta. Angelo’s specialities, apart from entrepreneurship, are search engine technology, intelligent transport systems and Artificial Intelligence.
Academically, Angelo published over 23 peer reviewed publications mainly in the UK and US, read for various degrees including a Ph.D. in Computer Science and AI and helped raise funding on various EU projects in the UK. Angelo has also won a bronze medal at the International Olympiad in Informatics (IOI) in 1995.
When not creating innovative products and technology, Angelo can often be seen running marathons, cross-country skiing, travelling and doing outdoor fun stuff.
Below is our interview with Angelo, which we hope you are going to enjoy as much as I did conducting it.
When was the first time you have started seeing the big picture in developing AI?
Angelo: I have been interested in AI from the very beginning of my career. I grew up watching a lot of science fiction movies that touched upon various aspects of AI since I was kid – Star Trek in particular was a major influence on my nascent thoughts. I was lucky that at the time I started learning Computer Science in depth, we had an influx of foreign educated lecturers who actually started an AI undergraduate course in Malta, where I read for my first degree. With hindsight, this was not so common in 1996 and I was lucky to have had a fairly early exposure to AI techniques. Some of the people originally involved in the first generation of AI in Europe, who were based in Switzerland and the UK, had also moved to Malta at the time. I programmed my first neural network in 1998 – and quite frankly, the main factors that have changed, apart from minor improvements to the algorithms themselves, are the massively increased availability of huge quantities of data and the hardware advances that has made fast computation of results possible. The big picture was immediately obvious even back then, but in the late 90’s the application of AI was rather limited due to the processing and data constraints. Over the last twenty years I had applied elements of AI steadily in the Intelligent Transport Systems, Gaming, Retail and Data Analytics industries. Recently, after exiting from most of my operational roles in the companies I had setup, I spent almost half a year thinking about AI and the future. I have finally come to a concrete vision of how the AI of the future will look like – but you could say that the journey to get here has taken twenty years already.
I know that readers usually conduct researches and learn what AI is. However, could you give us a glimpse into how an AI project is actually developed and what are the key elements of setting it up?
Angelo: The key elements in setting up an AI project is to understand what unique data you can initially collect to train a useful system, then building up a core of initial users, and using the growing community to generate even more data that makes the system even more useful, in an ever-increasing virtuous circle. AI specific system configuration and a proper design of the architecture, data flows, scaleable storage systems, and so on are also needed. Naturally, the idea has to be sound and commercially viable as with any other project. The data and platform architecture emphasis is more key in AI projects than for other more conventional approaches.
Besides the actual result of an AI project and the vision behind it, does it involve passion of the programmer or the person behind the project? If so, can you tell us what you are working on currently?
Angelo: Current technology requires the AI system to be trained and tweaked rather extensively, especially in the beginning, when different parameters and configurations of the architecture are being considered. The people involved need to have knowledge of the problem domain being solved and as much in-depth knowledge possible of how the AI solutions being used actually work. At the moment, I am identifying different verticals in which AI can be applied. The underlying vision is to develop AI that works in an ethical manner, that impacts society positively and that assists – rather than replaces – people. I need to clarify that the current solutions for AI focus on a specific application area (something called narrow AI). We are quite far away from having a more general purpose AI system (something called AGI – for Artificial General Intelligence). My current focus is purely on narrow AI solutions, with a step-by-step vision of building up something leading to AGI over the course of many years. The current focus is purely commercial in nature – finding practical application areas that will generate revenue and help free up people’s time in order for them to achieve their full potential. We are working on assistive AI solutions starting from the area of corporate services and have already rolled out a successful solution in Latin America that uses AI to generate large scale predictive traffic models in real-time – something beyond the capability of what people can ever do.
A few months ago, you said in an interview for lovinmalta.com: “Imagine a future where AI can create living content which is so flexible that each individual user reads differing versions of the same article based on their social media patterns and receptiveness to the subject matter. This, like most things AI related, is incredible but also incredibly scary. This is why a code of ethics and strong legal guidelines are fundamental.”
When developing AI, how many companies actually consider keeping ethics above manipulation?
Angelo: Unfortunately, a lot of companies do not consider ethics from day one when building AI solutions. The recent case involving Google and Project Maven is a case in point. Luckily, in that case, employees took matters in their own hands and forced the company to rethink the strategy. What we are doing is to have ethics at the heart of our solutions from day one. To this effect we have on our advisory board the United Nations expert on ethical AI, Prof. Noel Sharkey – who many of you may know as the main judge in the popular series Robot Wars – who will be helping us developing our code of ethics both for our employees and also for our AI products. Additionally, our chief legal expert, Dr. Olga Finkel, is working on an ethical framework that should apply to future AI products in all shapes and forms, basing upon her experience in drawing up legal frameworks and her own past expertise in computer science.
In an ideal world, AI can assist us in our social life and could keep us in a “positive mood” by filtering the “negative” information we could potentially receive on a daily base. This could be implemented in our mobile devices and monitor our social media activities. Do you think that this would be an ethical use of AI? Do you see this as a possible cure for narcissism?
Angelo: I think that AI solutions should impact society in a positive manner and at a minimum should never knowingly harm people. I do think that an AI assistant that filters out negative information would be an ethical use of the technology – although there are obviously cases where you do need to know negative information not just positive, for example, in the case of bad non-fake news that is relevant. AI can help in having people achieve a better balance in their use of social media and reduce excessive dependence and addiction to it. In my recent past, we had worked on various solutions for the gaming industry that identifies possible addiction – I believe that similar solutions for social media addiction would be useful in weaning off people from wasting too much time on their smartphones and instead spending it in creating meaningful relationships with people.
I’ve read somewhere that your profile is a realization of the mantra by which you work by ‘To create scalable, repeatable technology businesses that lead by innovation’. Do you think that a fully autonomous AI running all aspects of our lives is the final frontier?
Angelo: I think that we will increasingly delegate more parts of our daily lives to AI as the technology gains more capability and as it earns our trust by behaving in expected and socially acceptable manners. I don’t believe that people will ever completely relinquish control over key aspects of their lives to AI – and nor should they in my personal opinion. Some key decision making should always be left in human hands, possibly with AI helping presenting a clear view of all the available options to assist in making the best decision. However I do see that in the future, AI will be as pervasive and ubiquitous like electricity is nowadays.
For the final question, I would like to address a stereotype one. Should people worry about AI or embrace it as they embraces any other tech that came along the way?
Angelo: AI done right will not be just any other tech but will impact the way we go about our lives – my favourite analogy is that it will be a society-changing event like the invention of the steam engine, electrical power and solid state electronics. I believe that the variant of AI that I keep pushing forward – that is, AI done in an ethical manner and that impacts society positively – should be wholly embraced by people as it will bring revolutionary positive progress in human society.
If you you have a topic you would like to share with our audience, we are ready to have an interview featured with you. Make sure you send an e-mail to [email protected] and we will get back to you to set it up!