Let’s Take Back our AI Future – conversation with AI expert, George TileschGeorge Tilesch, an internationally renowned expert of Artificial Intelligence, was the online guest of the Institute of Informatics of Corvinus University at the start of May.
George Tilesch, an internationally renowned expert of Artificial Intelligence, was the online guest of the Institute of Informatics of Corvinus University at the start of May.
George Tilesch, is an internationally renowned expert of Artificial Intelligence, member of the AI Ethics Working Group and advisor for the White House, Founding President of PHI Institute for Augmented Intelligence, consultant with government and corporate leaders worldwide and AI author. He lives in Silicon Valley, but recently he arrived in Budapest since his new book “Between Brains: Taking Back our AI Future”, co-authored with Dr. Omar Hatamleh, former Chief Innovation Officer of NASA, was published in Hungarian by Libri. On the 3rd of May, at the invitation of the Institute of Informatics he was the guest of Corvinus University in the frame of an online conversation.
Andrea Kő, the director of the Institute of Informatics and the host of the event recalled that George Tilesch had worked with Microsoft and NASA, advised the World Economic Forum, as well as various companies and governments. He is the AI Ambassador of Neumann Society. AI has already transformed our daily life and to understand it requires a truly multidisciplinary approach, as the book emphasizes – recalled Andrea Kő. From a university point of view, it is of special importance to foresee how AI affects and reshapes higher education. Due to the ongoing revolution in education, also referenced in the book, it is important to clarify what are the changes caused by AI. What are the challenges that our system of education and its professors have to tackle? What are the new tendencies at the firms, startups, in the business models themselves that our students have to be prepared for? AI in the higher education raises methodological issues as well as questions about the content on AI to be provided. Due to accelerated technological development it is difficult to identify which knowledge can be valid a few years from now when our students try to use it in their actual jobs.
The content of the various e-learning courses has to be rapidly upgraded, often from non-academic sources, in order to keep up with the accelerating technology development – pointed out George Tilesch. The role of the professor will be to elaborate on use cases and case studies during seminars in order to discuss the practicalities of AI implementation in daily work. In the US, Tech companies are much more interested in one’s aptitude and industrial certificates than in a university degree. The universities have to provide the basics, the context and the capacity of independent thinking, while through marketable courses they need to integrate up-to-date content from industrial and research players. It is crucial to develop primarily the students’ soft skills. Recent WEF surveys indicate that creativity, empathy, the ability to cooperate, the readiness to innovate are among top skills in 2021 and beyond. The knowledge of technology is a general requirement. An important mental shift is accepting the necessity of continuously developing new skills required to fulfil ever-shifting new jobs.
There is a high interest toward AI and its broad transformative effects as it was evidenced by the recently published Hungarian edition “BetweenBrains”, that turned out to become the fifth most popular book – pointed out George Tilesch. He recalled that the trigger for the book was originated from a series of conversations in 2016 at the premises of the NASA among representatives of the big tech firms and the US Government. The technology experts publicly shared optimistic views regarding the bright future of the AI. However, during the coffee breaks the very same executives admitted to having significant concerns regarding the AI related technologies that their companies generate with robust investments.
It was a final trigger to George and his co-author to bring about a truly interdisciplinary approach into the AI related conversations. It became clear that while the AI affects the life of everybody, its development is driven by a rather narrow circle of decision makers. It makes it urgent to engage the people, especially responsible thinkers in a broad conversation, in order to avoid a rather dangerous course. “BetweenBrains” aims to inform digitally literate citizens to sort out the distortions from the media, knowing that the terminology makes it a more difficult task to achieve. The aim was enabling people to shape an AI-driven future instead of becoming passive, voice innocent bystanders of AI-induced transformations.
How can we describe a well-functioning AI paradigm that will genuinely serve the best interest of the humankind? – asked the co-moderator of the event, József Veress, assistant professor of the Institute of Informatics.
Very consciously, we prefer to talk in our book about augmented intelligence rather than artificial intelligence, as the way to go – pointed out George Tilesch. Augmented intelligence means public interest technologies capable of improving the human condition. These can facilitate the empowerment if human beings rather than replacing them with intelligent machines. In the book we argue for the necessity to reconnect technology development with strong ethical requirements both in the business and in the government sectors. The moral fundaments and the trust driven approach are of crucial importance for the success of AI, especially augmented intelligence related developments. It is a serious warning signal that only 20 % of the students of Technology Universities are studying Ethics or at least have such a subject in their curricula. Technology, Ethics, social and cultural issues should be handled simultaneously and from all humankind’s point of view in such constellation, AI is simultaneously a challenger and an enabler.
The “AI for good” approach aims to facilitate social and environmental improvements similarly to the UN’s 17 Sustainable Development Goals. As the 2017 study of McKinsey indicated, there are broad overlaps among these positive scenarios what provides the technical basis to avoid frequently cited dystopias. The Green Digitalization, the conscious connection of the efforts to rely on selected patterns of digitalization to tackle the climate change becomes growingly important as a requirement. AI provides very effective tools to model and simulate climate change related processes and identify due changes. The tech companies are reluctant to talk about the fact that the emissions generated during the preparation of a simple model of machine learning equals the emissions of five cars during their entire lifetime. The large companies already have to tackle the environmental aspects. For the smaller firms the new business models can offer a genuine alternative and also the advantage of the first comers. Not by chance the EU has launched a program aiming to connect AI and the new Green Deal policies where George Tilesch was invited together with another famous US expert.
András Gábor in his intervention recalled that the AI hype of the 80s – followed by an AI winter – was characterized with a triple approach embracing questions like: How does human thinking take place? How can a problem be solved? How tools and algorithms facilitate problem solving? Currently when AI is tackled in a business and / or public service context the first two components seem to raise less attention. Nevertheless, it is important to handle all these three fields simultaneously following a complex manner instead of narrowing down everything to the machine learning.
Such complex approach is inevitable in case we want to achieve a utopic rather than a dystopic outcome – emphasized also George Tilesch. The current “summer” is the outcome of hardware, software and data constellations where profitability and scalability can be achieved. However, an AI winter can re-emerge again if excessive expectations appear. Many researchers are working on such well-known tendencies worth to be reactivated such as Symbolic AI.
In front of the robustness and transformative capacity of AI driven marketing tendencies we have to rebuild the systems of checks and balances at individual as well as at a societal level. AI is used primarily as a marketing tool, and it remains unnoticed while going under our skin. Since it knows our habits much better than we do it also enables to monetize them therefore the individual has to rearrange its “mental immune system”. The current open-ended constellation creates a new situation where rules are changing and we have to rethink our thought, institutional and social patterns.
Professor László Zsolnai indicated that a major lesson of the last 30-40 years of business ethics is that without renewing the business models there is no chance to consider environmental and ethical requirements in business at the level of core values. The Ethics of AI can become a real issue if new business models emerge.
Indeed, new business models are of an absolute necessity – agreed George Tilesch. The incomprehension around Ethical aspects is mind-boggling. Even within the US government the conflict emerges with those technical experts who try to dramatically water down Ethics by delimiting it in a way when human beings are excluded from its originators, staying it as a merely technical problem. In fact, they are taking away the substance of Ethics. Another extreme is the swarm of startups who write software to handle Ethical issues. There were few more destructive developments in the economic history than the current models of the data economy or economy of algorithms. The current accidental data monopoly of e.g. Facebook and Google should be changed. In fact, it is of crucial importance to erect new business models. A fundamental upgrading of the rules has to take place. Hopefully the new generation of the Western youth will take the lead who are data conscious and who pick a job in a company depending on its Ethics. There are studies indicating that members of the generation Z are rather value conscious while selecting a job.
The current powerful data systems can be dangerous since the related quasi-intelligence is much more powerful than intelligent. Our current pilgrimage towards a general AI from the current narrow AI can become a rather dangerous period. The recently fired member of the Google AI ethics team has raised the issue whether the efforts for AI assistants to better understand human conversation can reach a volume and quality of data at an uncontrolled manner. The outcome can be robust and powerful – although not intelligent.
The emergence of superintelligence was mostly foreseen as an incidence, also in science fiction novels. Looking at the mainstream tendencies and having a peek at behind-the-scenes processes, currently there are no signs of the emergence of self-conscious machines. But the fact that our lives are increasingly driven by huge data collecting and processing systems is definitely worth to be concerned…
Currently we have very few if any conjecture on what takes place in the virtual space and don’t have the political capacity to demilitarize and de-commodify the AI. Although there is an ongoing UN dialogue on preventing autonomous weaponry to be used four members, including China, Russia, the USA, Israel, don’t support the process. No one can seriously count on a timely SALT agreement, either no deal can be reached, or a completely meaningless paper can become the outcome. Very big catastrophes must take place to change this trend. Unfortunately, no direct connection between positive impacts of the recognition of the environmental hard barriers and a genuine control ensuring peaceful AI implementation can be discovered today. The positive signals in current high-level politics about the readiness to cooperate on climate issues can be encouraging in the long run. But AI travels exponentially. In the context of AI, serious scandals can raise due awareness and launch genuine preventative steps in regulation and public awareness. The topics that can launch due changes probably are data consciousness, changes in the world of work, as well as distribution-related issues. AI is very much facilitating the concentration of power and profit and mass societal movements are required to enforce due changes. Let’s hope to keep our sobriety and sanity.