Summary of Professor Richard Susskind OBE KC (Hon) - The Lionel Cohen Lecture 2023, The Hebrew University

This is an AI generated summary. There may be inaccuracies.
Summarize another video · Purchase summarize.tech Premium

00:00:00 - 01:00:00

Professor Richard Susskind, a global thought leader on technology in the law, gave the 2023 Lionel Cohen Lecture at The Hebrew University on the unique relationship between the British and Israeli judiciaries and Artificial Intelligence (AI) in the law. He stressed the need for legal professionals to adopt a different mindset towards the use of technology and distinguish between automation and innovation. Susskind predicts a few law tech startups will emerge and fundamentally change the legal landscape by 2030, and while AI will play a fundamental role in the transformation of law and legal institutions, it won't move at that pace due to cultural, regulatory, and commercial reasons. He warns that machines will be more capable than humans in almost all respects and will redefine what it means to be human and work.

00:00:00 In this section, the speaker, Professor Richard Susskind, is introduced as the speaker of the 2023 Lionel Cohen lecture which focuses on the unique relationship between the British and Israeli judiciaries, particularly on the topic of artificial intelligence and the law. The lecture has been delivered since 1953 by leading jurists from the UK or Commonwealth countries and is regarded as the Faculty of Law's most important and prestigious annual lecture. Professor Susskind is a global thought leader on technology in the law and gives insights on the mindset that we need to have in the legal world in thinking about the future. He also suggests that there are three main drivers for change in the law, one of which is technology, which is currently affecting the way we live.
00:05:00 In this section, Professor Richard Susskind speaks about the importance of mindset and technology in the legal world. He highlights the need for a different attitude towards the use of technology, asking legal professionals to think about the fundamental value they bring to their clients. He distinguishes between automation and innovation, stating that the game-changing uses of technology are the ones that allow us to do things that previously weren't possible rather than just automating current processes. He argues that we need radically new thinking, enabled by technology, to solve global issues such as the access to justice problem.
00:10:00 In this section of the lecture, Professor Richard Susskind discusses the rapidly advancing technologies that are driving change in the legal world. He notes that cost competition and technology itself are key drivers of change, with legal services becoming unaffordable for almost everyone. He also highlights the emergence of new players in the legal ecosystem, including the big four accounting firms and legal technology startups, which are disrupting the traditional models of legal service delivery by offering cheaper, better, and quicker legal services. The competition that these new players bring is good for clients and existing law firms, encouraging modernization and improvement. Susskind predicts that a few of these law tech startups will emerge and fundamentally change the legal landscape.
00:15:00 In this section, Professor Richard Susskind highlights the increasing capability of computer systems to take on tasks that were once thought to be undertaken only by human beings. He focuses on the explosive growth of underpinning technologies in the world today. With the exponential growth of neural networks, the performance of machine learning systems is doubling every 3.5 months, meaning that over the next six years, it will increase 300,000 fold. The technology is advancing so quickly that we have no sense of what is coming next, and there is no finish line in sight.
00:20:00 In this section, Professor Richard Susskind talks about the unprecedented technological advances that are happening on our watch and how it will lead to a transformed planet in a few hundred years. He believes that by 2030, our lives will have been completely transformed by technologies that have not yet been invented, and this is both exhilarating and terrifying. He also discusses the impact of AI on the legal landscape and believes that almost all short-term predictions are hugely overstated, while almost all long-term ones are hugely understated. He predicts that AI will play a fundamental role in the transformation of law and legal institutions by 2030 but acknowledges that it won't move at that pace due to cultural, regulatory, and commercial reasons.
00:25:00 In this section, Professor Richard Susskind discusses his early work in AI and law, which began with his undergraduate thesis at Glasgow University in 1981. He later did his doctorate on AI Law at Balliol College in Oxford, where he explored the limitations and implications of computer systems that could solve legal problems. He designed a system based on the consensus in the field of analytical jurisprudence and identified the limitations of the systems in mind. His area of application was family law, and he identified that the systems could solve problems that legal experts found easy, but the rest of us found hopelessly difficult. In the late 1980s, Susskind worked with a group of scholars and friends on the first generation of AI in law systems, which were rule-based expert systems. They were thinking about how to move this research from the laboratory into the marketplace.
00:30:00 In this section, Professor Richard Susskind talks about the first generation of AI, which involved expert systems. He explains that these systems were developed by hand-crafting decision trees, or complex flow charts, based on the knowledge of human experts. Although the intelligence of the system lay more in the pre-processing rather than in the system itself, it was still considered a breakthrough in the late 80s. The system, which was designed to help lawyers navigate cases in areas such as negligence liabilities, product liability, personal injuries, and fraud, surprisingly often came up with answers that even the expert who designed it did not contemplate. The system also offered an explanation of its lines of reasoning and told the user all the fundamental dates on which it relied.
00:35:00 In this section, Professor Richard Susskind explains the challenges of developing AI systems for the legal industry. He notes that while technical feasibility was demonstrated in the late 80s, commercial needs have not driven their development due to their high cost. Additionally, the advent of the web, Brute Force Computing, and machine learning were not anticipated back then. However, as computer processing power increased, Brute Force Computing became a new way of operating for AI systems. Furthermore, rules-based expert systems gave way to systems that learn on their own, leading to more sophisticated AI.
00:40:00 In this section, Professor Richard Susskind explains how machine learning systems work by using an analogy with how humans learn languages. He goes on to discuss second-generation AI in law, which specializes in predictive and generative tasks. He talks about Lex Machina, a predictive technology developed at Stanford University, which is capable of predicting the outcome of patent disputes with remarkable accuracy, using data about past cases instead of the legal method. Professor Susskind believes that if computational statistics or AI can deliver quicker, cheaper, and more accurate solutions to a specific class of legal problems, the market will demand it over traditional legal methods and lawyers.
00:45:00 In this section, Professor Richard Susskind discusses the capabilities of AI in the legal space. He mentions how predictive AI systems are being used for document review and litigation and how a system is more accurate than a junior lawyer for due diligence. He also brings up generative AI, a new development in the AI space, and how it has quickly become adopted by over 100 million users. Susskind's personal experience with GPT, a computer system that can create coherent written text, has convinced him that it is the most remarkable development he has seen in 40 years of AI. He believes that most computer scientists underestimated the capabilities of AI and that they made a false distinction between narrow AI and general purpose intelligence.
00:50:00 In this section, Professor Richard Susskind discusses the capabilities of chat GPT, an AI system that can mimic human conversation to a remarkable degree. He highlights how this technology is becoming increasingly capable, and how it can already perform tasks such as summarizing and drafting contracts, supporting judges in preparing their decisions, and conducting legal research. Professor Susskind emphasizes that these systems do not need to replicate human thinking processes to perform at a high level. He warns against the mistaken assumption that true AI requires replicating the thinking of human specialists and argues that this is the folly of the architecturalist view of AI.
00:55:00 In this section, Professor Richard Susskind discusses the role of judgment in the legal profession and why it's a fundamental problem. He argues that clients do not ask for judgment, but for solutions to their problems of uncertainty. Machines are much better at handling uncertainty, and lawyers will need to build or compete with emerging systems that will replace many traditional legal practices. The next decade is not a time of unemployment but a time of redeployment where lawyers need to become the people designing these systems that solve clients' legal problems. He warns that machines will be more capable than humans in almost all respects and will redefine what it means to be human and work.

01:00:00 - 01:30:00

In a lecture at The Hebrew University, Professor Richard Susskind discusses the potential of technology to become embedded within humans, leading to transhumanism, and the challenges this poses in terms of power and ethical boundaries. He believes that AI will transform the legal profession beyond task substitution, leading to new ways of solving problems through technology and the potential for incorporating a wider range of discourse into judicial decision-making. Susskind emphasizes the importance of regulating AI development, ethical and philosophical discussions, and deeper considerations of what is worth preserving in the legal system, rather than surface features. Additionally, he raises concerns about unregulated AI being concentrated in companies and the potential for machines to convict people before committing a crime, emphasizing the need for more dialog and critical decision-making.

01:00:00 In this section, Professor Richard Susskind discusses the potential for technology to become embedded within humans, leading to a new era of transhumanism. He highlights concerns about the concentration of power and wealth in the hands of a small number of companies delivering these systems, and the lack of clear ethical boundaries being considered for advanced systems. Professor Susskind suggests that AI will transform the legal profession beyond task substitution, leading to a fading of lawyers like craftsmen, and new ways of solving problems through technology.
01:05:00 In this section, Professor Richard Susskind discusses how technology will change the way we practice law and potentially challenge legal institutions such as the institution of contract. He believes that much of the law created in the future will be immediately embedded in our buildings, working processes, and lives. He maintains that principles of the rule of law can be maintained as long as these systems are built to comply with it. However, he has concerns about distributive justice and the emergency systems that could overcome the access to justice problem but might be struck down on grounds of other principles of justice. Professor Susskind reminds his audience that the question of what is the future of lawyers is the wrong question; instead, we should be focusing on how we will solve legal problems in the future.
01:10:00 In this section of the lecture, Professor Richard Susskind addresses the question of whether AI can solve legal problems better than lawyers. He argues that the answer is yes, and that AI has the potential to handle certain categories of disputes more efficiently and cost-effectively. However, he acknowledges that there is still value in thinking about legal questions and participating in legal discourse, as it plays a crucial role in constructing normative and moral discourse. He suggests that machine learning systems could be developed based on the discourse and debate that is fundamental to the judicial process, but still acknowledges the importance of hearing and discussing hard cases.
01:15:00 In this section, Professor Richard Susskind discusses the possibility of incorporating a far wider range of discourse into judicial decision-making through the use of machines. He believes that public discourse could be more extensive, richer, and consultative, creating arguments, propositions, and conclusions that have intrinsic value that may not necessarily lead to decision-making. Although machines may not generate a richness or subtlety of argument that surpasses human beings immediately, he acknowledges that it is conceivable in the future. Therefore, he contends that modern jurisprudence ought to explore these kinds of issues.
01:20:00 In this section, Professor Richard Susskind emphasizes the importance of having a deeper understanding of what is worth preserving in the legal system for its own sake, rather than focusing on surface features such as whether or not judges wear wigs. He challenges the assumption that courts must be physical places where people congregate to resolve differences, arguing that a court service that is only available to a tiny minority of people is like "having a Rolls Royce for one and everyone else is walking." He emphasizes that AI development is too important to be left solely to AI scientists, and that regulation, ethics, and philosophical discussions are essential to shaping the future of society and humanity.
01:25:00 In this section, Professor Richard Susskind expresses his concerns about regulating generative AI too early, as he believes that we need to take time to think through and test this technology. However, he also acknowledges the dangers of unregulated AI being concentrated in companies that are changing our lives. Additionally, in the legal context, Susskind and David Nelken discuss the possibility of future AI systems being too perfect to the point that they predict criminal behavior, raising concerns about the fairness of the judgment process. Susskind argues that legal academics should put themselves in the shoes of non-lawyers and be aware of how they might view the judicial process. However, he also acknowledges that AI technology might eventually advance to the point where people will prefer a system than a human judge.
01:30:00 In this section, Professor Richard Susskind stresses the need to consider the ethical and philosophical implications of machines that are more capable of identifying flaws and shortcomings, including crimes. He notes that machines could potentially be used to convict people before they have even committed a criminal fault, a future that he finds disconcerting. While he acknowledges that machines might help us think these issues through, he believes that a lot more dialog is necessary to make such critical decisions.