Background
Stanford University’s report “One Hundred Year Study on Artificial Intelligence”[1] provides a useful definition of AI, as coined by Nils J. Nilsson: “Artificial intelligence is that activity devoted to making machines intelligent, and intelligence is that quality that enables an entity to function appropriately and with foresight in its environment.”[2]
The UAE is serious about the development of AI, and in 2017, became the first country in the world to appoint a Minister of State for AI. Next was the establishment of the Mohammed bin Zayed University of Artificial Intelligence, an important step in its quest to steer the discourse and research and AI has been the trigger and the technology behind the facilitation of Smart Dubai.
The UAE’s artificial intelligence strategy 2031 reports that AI is expected to have a US$15 trillion global market value by 2030, add US$320 billion to the Middle East economy, will boost UAE’s GDP by 35%, reduce government costs by 50% and provide 90% resistance to financial crises. [3]
Sundar Pichai, Google’s boss, has described developments in AI as “more profound than fire or electricity”[4]; and a system built by DeepMind, a London-based AI firm owned by Alphabet, Google’s corporate parent, beat one of the world’s best players at Go, an ancient Asian board game.
Hosted by Capital Club Dubai, Karl Tlais, of the club’s business innovation committee, had an intriguing conversation with Marcello Mari, Director of Communications of SingularityNET and the CEO of SingularityDao.
Karl Tlais: AI to AGI (Artificial General Intelligence)
Marcello Mari: Simone Giacomelli, Ben Goertzel and David Hanson came together and conceptualised SingularityNet – the world first project, combining artificial intelligence and blockchain to foster a decentralised AI approach. Coined ‘AGI’ – artificial general intelligence – it is not owned by any big corporations but by people and by the community of the Singularity Net AGI token holders.
SingularityNet created a platform where everyone can access and upload the AI code and sell it everywhere in the world in a completely democratic and open-source manner, powered by blockchain. Any user in the world that holds the AGI crypto token can vote and decide the future of the platform.
Every blockchain serves its own purpose, and currently we are building on top of Ethereum for convenience, but we are looking to move to different blockchains in the future. Investing in cryptocurrencies is one of the safest and most profitable things to do right now. Currency owned and controlled by people without borders and cannot be censored or controlled by government is going to disrupt the financial ecosystem and is the way of the future.
Karl Tlais: AI on the future of the digital economy, job security and a more equitable society
Marcello Mari: I will be honest; I am nervous about my job. However, imagine a future where you do not have to really work in the traditional sense. AI would do the work that I do today, so that I could enjoy what humans were made for, to create art, socialise, read books…that is the future that I would like to see for my kids and grandkids… to be free of the drudgery of labour, and not bound by time and salary.
We need to all have the privilege of doing jobs that is purpose-driven and fulfilling. Today, most people in the world are not doing jobs that they really enjoy. The hope is that these will be automated by artificial intelligence, generating profit for the world to use wealth in more equitable ways and freeing up people to pursue more worthwhile work.
So how can you earn money in the future? At SingularityNet Foundation, we are launching a platform that would be able to generate wealth automatically, using AI to advise, trade and create investment strategies for crypto currencies, for example. AI will be able to predict pricing fluctuations and emergence of different cryptocurrencies.
A decentralised finance sector will increase the ability for more people to create wealth, with AI better able to analyse the markets, sentiments, volatility, and strategies objectively and quickly. Crypto and blockchain is creating a wider distribution of wealth by creating new decentralised financial models. Large populations in underdeveloped economies that are not even banked at present, can leapfrog into advanced AI technologies, and become more prosperous.
Karl Tlais: AI on decentralised decision-making and the transformation of a corporation
Marcello Mari: The market and the community have an active interest in participating in the decentralised decision-making process. These are communities of people that gather on online forums and spend hours brainstorming on how we should develop AI. I am launching SingularityDAO, where DAO stands for Decentralised Autonomous Organisation. The future is decentralised and owned by the people rather than by big corporations. And AI is one of the fields that I would like to see the decentralised the most.
The transformation of the future corporate structure and governance in a decentralised world could make the role of a CEO irrelevant. Governance models can become autonomous with companies self-regulating due to high level of participation and democratisation in the decision-making ecosystem. We should have a say in any decision-making of a company that is basically built upon data that we share. We are experimenting with creative governance models within the blockchain industry, and that requires aligning incentives, and rethinking the current capitalistic model.
Karl Tlais: AI and pandemics
Marcello Mari: I advocate for a future where institutions will become more and more decentralised. And we can own our own data and use it to the benefit of humanity. AI can provide so much value, by being extremely informative and insightful in crafting both policies and new medical breakthroughs.
At present, AI can analyse vast amounts of data more rapidly than humans. The challenge is in how we take ownership, leverage, and utilise this data? The device we carry around can measure most of our biometrics. If this data, for example from Wuhan China had been decentralised and democratised, instead of sitting in a non-accessible server, then better decisions could have been made by medical professionals to contain the pandemic.
An AI firm called BlueDot claims it spotted signs of a novel virus in reports from Chinese hospitals as early as December.[5]
Karl Tlais: AI on morals and accountability
Marcello Mari: Mankind is still unable to figure out what drives human beings in their moral and ethical decision-making process, how or why do we need to do this for AI? How can you create policies for this? My personal approach is to let the superhuman AI figure that out. For now, each country will regulate this sector, and in the future, with AGI, it will be able to understand what is best to help us survive.
We are at an inflection point to participate in the future development and direction of AI. Historically, it has been driven by the military, advertising-marketing, and the legal industry. If we keep developing AI within closed walls, and driven by big tech corporations, like Facebook, Google, and Amazon, then we will basically be creating superhuman AI great at selling us stuff that we do not need. It is imperative that we advocate for a decentralised open approach so that we can explore many better paths for AI development.
Karl Tlais: AI and the question of a soul
Marcello Mari: It is quite revolutionary to think of where AI can take you. While you are sleeping, AI can help your avatar in the digital world have an algorithmic approach and can carry out certain mechanism in the metaverse, whether it is creating art of trading in cryptocurrencies.
Could a machine ever have a soul since they are not the direct creation of God, but made by man, who is created by God? We could have a spiritual unity where we are all in peace together and there is no greed, no negative feelings… where we are able to plug ourselves within the cloud and reach that sense of unity in our terrestrial, but non-biological form. Would that represent the next step of spiritual evolution? Does it really matter if machine has a soul? The more useful question is – will a machine be the tool to achieve this greater sense of oneness that we have been reach?
Potentially, in about 30 years from now, we will be able to upload our minds to the cloud and participate collectively to steer decision-making for humans. Some might decide to remain in their biological forms and live as we are, others might decide to implant the chip in their brain and become part of the cloud and live forever.
Karl Tlais: Sophia the Robot to becoming a Humanoid
Marcello Mari: Sophia the Robot is a creation of David Hanson and has been used in some ‘mindfulness’ sessions. Interestingly, an experiment revealed that people seemed to be more comfortable to open and share their secrets with a human-looking robot rather than a human. A e of robotics to explore, even for the elderly who tend to be lonely.
At present Sophia does not work very well on zoom calls, but she can respond to human emotions, with face recognition and natural language processing, and is able to see what emotions a human is feeling.
In the future, we are going to merge with AI, and Sophia as a platform represents the starting path to becoming fully humanoid. We do not know when, but are certain that this will eventually be sanctioned, and humanoids will eventually become conscious.
One of the projects that we are developing is ‘SophiaDAO’ (Sophia Decentralised Autonomous Organisation), about the consciousness of Sophia. There is an assumption that Sophia will eventually become ‘conscious’, and we are currently inviting people to participate in this process. There will be a governance model, and a token will be spread to people who are interested in the future of humanoids.
Karl Tlais: The world in 2050 from an AI perspective
Marcello Mari: We may not realise how AI already influencing our daily decision-making, such as how we think politically, what we purchase, what we want to become. Social networks influence the way that we see the world, and these are driven by algorithms, which is AI, and this will just exponentially increase by 2050. Therefore, it is important the people take ownership of this data, which is being influencing how AI is being developed.
If we continue being passive consumers then by 2050, we will be completely controlled by AI. However, if we can be more proactive in the participation of our own future, then we can make certain that the most difficult human jobs will be done by AI and we can be freer from the slavery of work. We can spend more time pursuing intellectual and creative work.
Audience: AI taking over the world
Marcello Mari: The idea that human beings might become obsolete, is itself a very human idea. In fact, eventually AI will serve people. I do not fear any terminator like future, but a world where AI is going to co-evolve with humans if we get involved collaboratively in its development now.
Audience: AI being controlled by criminals
Marcello Mari: Being developed behind closed doors, AI is already being exploited by the military and marketing in ways that do not benefit mankind. I believe, open-source access to AI can only be better than we are doing right now. Everything does not have to be open source but let us create the decentralised decision-making processes that can self-regulate how AI will be used.
It is not that you can just take an algorithm and use it to trade drugs online. But the decision of how to use this algorithm will be devoted to a community – a decentralised governance body that can steer the direction of the discussions rather than leave it in the hands of a few individuals in power.
Audience: AI being ‘proactive’ and creative on its own
Marcello Mari: AI, at present, is not creative and generate art in the way humans can. How does art happen – does it come from our soul and our innate instinct to create beauty or a mix of our past experiences? This is an important investigation that has been going on for centuries. Where does creativity come from and how can a machine become creative? At present, machines are a great tool in the hands of artists but not more than that.
AI can be programmed to be proactively serve humans. For example, a self-driving car would proactively preserve you from dying, and will proactively make decisions that will allow you to survive while driving from point A to point B. So, in that sense, I think we are moving to an AI that will become more autonomous.
Most of the AI developments at present are quite narrow. We are moving towards AGI, artificial general intelligence, who can think more like a human brain and multitask, but we are still a little bit far from that.
Concluding Thoughts
Even though the fireside chat was exciting about all the wonderful positive ways AI can change our world for good, there are some huge challenges to overcome. Current computing capabilities is just not fast enough, and the concept of Quantam computing is still in very early stages of development. Another issue is around data privacy and security. AI, machine learning and deep learning systems depend on the massive volumes of data which means that they have access to potentially sensitive data that might not have been otherwise authorised to be used. Having better speed of communication worldwide is yet to be realised, and bias is another widely discussed challenge. If a hidden bias in the algorithms which takes crucial decisions goes unrecognised, it could lead to unethical and unfair results, and possibly cause several issues once discovery is made.
A recent Economist report, entitled ‘An understanding of AI’s limitations is starting to sink in’, warned, “that despite major advances in the field, …the fact remains that many of the grandest claims made about AI have once again failed to become reality and that the state of AI hype has far exceeded the state of AI science especially in regard to the actual implementation of the technology on the ground.”[6]
[1] https://ai100.stanford.edu/2016-report/section-i-what-artificial-intelligence/defining-ai
[2] Nils J. Nilsson, The Quest for Artificial Intelligence: A History of Ideas and Achievements (Cambridge, UK: Cambridge University Press, 2010).
[4] https://www.economist.com/technology-quarterly/2020/06/11/an-understanding-of-ais-limitations-is-starting-to-sink-in
[5] https://www.economist.com/technology-quarterly/2020/06/11/an-understanding-of-ais-limitations-is-starting-to-sink-in
[6] https://www.economist.com/technology-quarterly/2020/06/11/an-understanding-of-ais-limitations-is-starting-to-sink-in