Angelo Dalli is the CEO and Co-Founder at Umnai, a company that is creating new AI technology that can explain itself and its decisions with the objective of helping and assisting people rather than replacing them. As Umnai’s Chief Executive Officer, Mr Dalli is responsible for leading the company’s commercial, product, and technology strategy.
Mr. Dalli is also a speaker at the Malta AI & Blockchain Summit conference, happening on May 22 – 25 in Malta.
In this interview, Mr. Dalli shares insights on explainable AI, his company’s mission to develop AI technology that assists people rather than replacing them, combining AI and blockchain, and more.
Q: What is your background and how did you start Umnai?
I have been a graduate in Artificial Intelligence for a long time – almost 20 years. Back then, AI was not the hot topic that it has become today – I have always believed it will be the future of technology – it just took a bit longer than expected to get to this stage where the software, hardware and data availability are all now sufficient enough for AI to stay. In my previous ventures and companies that I have started, we had already started applying some AI to different industries – yet the current wave of useful applications have been largely generated by the next generation of hardware that allows fast and efficient processing of the type needed by AI systems.
Umnai is my ninth start-up – so I took quite a long time deliberating what to focus on before choosing to untangle myself from most operational commitments I had at that time and focus entirely on AI. The ideas for Umnai germinated over a long series of treks up in the Himalayas in Nepal – the peaceful atmosphere allowing me to think clearly and strategically to come up with a very solid and long term plan for AI innovation.
We are focusing on AI that can work together with humans and enter into an interactive dialogue with them, ensuring that people can examine the results of the AI and correct it, just like humans mentor children and other humans.
Q: Umnai’s vision is to “develop the next generation Artificial Intelligence technology that aims to assist, rather than replace people”. Can you elaborate more on that?
Umnai is creating new technology that has the potential to disrupt the AI industry completely, affecting so many different software and hardware verticals that the potential upside is very substantial. We are focusing on AI that can work together with humans and enter into an interactive dialogue with them, ensuring that people can examine the results of the AI and correct it, just like humans mentor children and other humans. In this way, people will work hand-in-hand with AI systems, using the AI to improve their productivity and freeing up precious time to be more creative and innovative. We are currently in the process of patenting our technology so we are rather limited in the amount of detail that we can reveal at the moment – watch out for announcements near the end of this year.
Q: What is explainable AI? Why is explainable AI important?
Explainable AI is AI that can explain the steps it has taken to come to an answer, together with the basis for its decision and any factors that have been used in coming up with those answers. Current narrow AI simply gives you an answer, without any explanation – this behaviour and lack of transparency can lead to a backlash against AI and also will lead to issues of trustworthiness as no one in their right mind would blindly follow an AI system’s recommendations without having a proper explanation and insight into the process that came up with such an answer. This is very important for regulated industries, such as healthcare, law, finance, gaming, and so on, where regulators world wide seem to be favouring the introduction of explainable AI in all regulated industries.
AI being able to explain itself is a key milestone if AI ever wants to be a collaborative tool or partner with humans, integrating better with human society – otherwise, it will be forever relegated to restricted and narrow applications that do not meet the public’s perception, needs and expectations of future AI systems.
Q: In which domains do you see the biggest impact for explainable AI in the coming years?
Explainable AI will firstly impact all regulated industries and also areas where there is a need for mission criticality and explanation. Anything that has an impact on human life (for example, driverless cars – that may cause accidents), or is required to make a significantly important decision will benefit from having explainable AI technology. The impact will be profound as people will get used to asking AI for explanations, and by following its logical processes learn when to trust its judgement or override it. AI systems are not as perfect as some hype makes them out to be – currently there is a lot of work needed to select the right data, make sure that the right features are selected and modelled, and also to train and re-train the model to keep it updated. Umnai’s solution allows for human knowledge to be directly blended with AI knowledge, allowing for the impact to be magnified significantly further.
AI being able to explain itself is a key milestone if AI ever wants to be a collaborative tool or partner with humans, integrating better with human society.
Join our weekly newsletter to receive:
- Latest articles & interviews
- AI events: updates, free passes and discount codes
- Opportunities to join AI Time Journal initiatives
Q: What is the AlphaStars initiative?
The Umnai AlphaStars initiative is to have a group of select early adopters who start using Umnai’s technology much before the general market, getting the early adopter benefits while in turn providing quality feedback that will shape the productization of the technology for different verticals and industries. We are slowly working with different partners and opening up the initiative to innovative companies who are completely committed to using AI as part of their core business and who have a valuable problem to solve using Umnai’s solutions.
Q: Tell us about the Umnai team.
The Umnai team has been hand-picked by myself to represent the core nucleus of the team that will eventually grow to a much larger team in the coming 2-3 years.
Virtually everyone in the team has known each other for more than 3 years, with some relationships going back more than 12 years. This is a big advantage in a world where startups tend to have a group of people who didn’t know each other that well and also goes against the present day concept of changing jobs every so often – we do not believe in such an attitude and my experience as a serial entrepreneur has showed me the importance of having a strong and loyal core in the team that enables the company to move forward fearlessly and at a high speed without friction.
The core team is supplemented by a world class advisory board having over 100 years of combined AI experience including three AI veterans: Noel Sharkey, Yorick Wilks and Mike Rosner. Yorick has won some of the highest accolades including the Loebner Prize in AI, the BCS Lovelace Medal, the Zampolli Prize in NLP, the ACL Lifetime Achievement Award amongst others. Noel is a well known British AI personality and is heavily involved in the UN efforts regarding ethical AI and the non-weaponization of robotics (you may have seen him as the head judge of the popular UK TV series Robot Wars!). Mike was one of the first person setting up AI research labs in Switzerland and is a key figure in Computational Linguistics and was one of the first people who started teaching AI in Malta.
Q: What are Umnai’s biggest achievements in the last 12 months?
Umnai’s biggest achievements have been in the breakthroughs we’ve made in explainable AI and related areas, solving problems that have been long outstanding for over 30 years. At the moment we are carefully testing and verifying our technology and in the process of patenting it in various countries around the world to build the foundations for what could be one of the most significant disruptions in AI that will make it much more useful.
Q: You will be a speaker at Malta AI & Blockchain Summit. What will your speech be about?
I will be speaking about the future of AI and how it will affect society – how will AI be integrated with people, how will it affect our jobs, education, way of working, what are the benefits and pitfalls, and so on. Also I will be addressing a lot of common concerns that I get asked about frequently in different events around the world.
Q: Where do you see the biggest potential for combining AI and Blockchain technology in the coming years?
I think that blockchain can be a useful tool for AI when it comes to saving a verified audit log of its decisions and explanations, and also to a lesser extent to serve as a useful payment method for API calls, etc. The main use of blockchain as a distributed ledger (DLT) is the main short term potential for AI though – the speed of the blockchain technology needs to improve though in order to be useful at a massive scale.
Join Mr. Dalli at Malta AI & Blockchain Summit
Mr. Dalli will be one of the speakers at Malta AI & Blockchain Summit happening on May 22 – 25 in Malta.
Angelo Dalli’s Speech
“DEMYSTIFYING AI: THE TRANSITION IS CLOSER THAN YOU THINK”
May 24th at 9:30