Intangiblia™
#1 Podcast on Goodpods - Intellectual Property Indie Podcasts
#3 Podcast on Goodpods - Intellectual Property Podcast
Plain talk about Intellectual Property. Podcast of Intangible Law™
Intangiblia™
Nicolas Torrent - Accessible by Design: How AI Can Open the Doors of Justice
Imagine opening your phone, describing your dispute in simple language, and getting a clear, data-backed path to resolution—without weeks of confusion or a wall of legalese. That’s the future we dig into with lawyer and legal tech builder Nicolas Torrent, who’s helped design online arbitration platforms and shape Switzerland’s legal tech ecosystem. Together we unpack how AI, user experience, and court data can turn access to justice from a maze into a map.
We start with the hard truths: price uncertainty, physical distance, and cognitive barriers keep people out of court. Nicolas lays out how legal design—plain language, smart workflows, and visual cues—can guide users step by step. Then we zoom into the power of data: aggregated outcomes that help people understand their odds, timelines, and likely costs, improving settlement decisions and restoring trust. Speed isn’t just convenience; it’s an economic catalyst. When fair rulings arrive sooner, families and small businesses can move forward with confidence.
We also explore a sustainable path. Nicolas outlines “profitable justice” that doesn’t hide rights behind paywalls: think low-cost online small-claims settlement tools that offer realistic ranges based on similar cases, with an option to escalate to a human judge. Pair this with supervised trainee reviews, pro bono, and targeted lawyer services, and you get a flexible market that meets people where they are. Along the way, we tackle big-picture risks—AGI race dynamics, quantum acceleration, and geopolitical stakes—and why open source, distributed authority, security, and personal accountability must anchor any public system.
Throughout, one principle stays constant: keep humans in control. AI should accelerate routine work, surface patterns, and translate complexity into clarity, while judges and lawyers apply judgment, empathy, and responsibility. If we design for inclusion, treat court data as a strategic public asset, and build with transparency, justice can become faster, fairer, and truly accessible. If this resonates, subscribe, share with a friend, and tell us: which part of the legal journey should be redesigned first?
Can't keep up with AI? We've got you. Everyday AI helps you keep up and get ahead.
Listen on: Apple Podcasts Spotify
Check out "Protection for the Inventive Mind" – available now on Amazon in print and Kindle formats.
AIs have an amazing ability to understand a person's emotional state.
SPEAKER_03:Okay.
SPEAKER_02:Um, and it overperforms humans. It outperforms humans. So being understood, the AI might very well give you the feeling that it understands you better.
SPEAKER_06:Hello and welcome to Intangible. Today we have a very special guest, and um we're gonna talk about uh the topic that is in everyone's mind, in everyone's mouth. So it's AI, but with a twist. We're gonna talk about accessible by design, how AI can open the doors of justice. So we have here Nicolas Torrent. Welcome to Intangibilia.
SPEAKER_02:Thanks. Thanks for welcoming me.
SPEAKER_06:He's a lawyer. Let me know if I'm correct or not.
SPEAKER_02:I don't know, it's true.
SPEAKER_06:For now, right? Okay, so Nicolas, or in Spanish, Nicolas. He's a lawyer who doesn't just follow legal innovation, he helps build it from designing platforms for online dispute resolution to shaping the legal tech ecosystem in Switzerland. He worked on online arbitration platforms, designed legal softwares to simplify contract management, and advocated for EU seen AI to automate and include. This conversation is about AI that is not a threat to the legal system, but a bridge that can connect courts to the people they're meant to serve. Let's explore how justice becomes scalable, human-centered, and even sustainable. So the first question, please introduce yourself and let us know how you came to be here in Switzerland and Geneva.
SPEAKER_02:Sure. So I was actually born here. So I studied here all my life. Um, I had a brief experience of a couple years in Paris to sell software. That was the arbitration platform you mentioned. And since then I came back to Switzerland and I've been building quite a lot of projects related to making the legal world easier to navigate for everybody, basically.
SPEAKER_06:Okay. And what, because usually lawyers we are quite traditional in the way we're trained, the way we are taught. What gave you this edge for going for technology or technological areas?
SPEAKER_02:Yeah, I guess, I guess um, I just like tech. I wasn't I was really bad at math, but I wasn't bad at language.
SPEAKER_07:Okay.
SPEAKER_02:So when it didn't work out with math, I tried doing polytechnics in Lausanne. But when that didn't work out, I thought, okay, well, I'll just go for words, it's logic as well.
SPEAKER_05:Yeah.
SPEAKER_02:Um, and that worked out better because there were very few maths involved, and many lawyers actually don't like math. So I feel that I fit right in.
SPEAKER_07:Yeah, yeah, yeah. It's it's I think it's a universal thing.
SPEAKER_06:There are some few tax lawyers who like math, but the rest of us uh is quite some, it's a it's a topic for another profession. Yeah, absolutely.
unknown:Yeah, yeah.
SPEAKER_06:And you've seen the legal system from inside the courtrooms and through the lens of legal text. In your view, what is the biggest obstacle today to making justice truly accessible for the general public?
SPEAKER_02:I think it's really a combination of elements. Um, there's no single factor because it depends who's looking at the problem. But the number one issue I'm hearing about is price. It's how much does it cost? Because people don't just think, oh, the court is gonna cost me that much. They also think I'm gonna have to pay for a lawyer. Um, and they have often misconceptions about how much a lawyer costs. So the price is an issue. It's difficult to navigate when you're alone and you're trying to figure out how much it's gonna cost you. People need predictability, they need budgeting, they need to be able to see things from their own uh economic situations perspective. Uh, another issue is uh the physical or the cognitive accessibility. If you're living outside of one of the big cities, it gets much harder to access a tribunal. And you might have some family obligations, you might have some professional obligations, you might have different kinds of constraints that prevent you from accessing a court. Um, and likewise, if you're a person with disabilities, it can also be quite a challenge to navigate a court system because it wasn't designed for people with disabilities. So this is again something that is not that hasn't followed the model the trends that we're seeing today. We're also seeing that the interface to operate in the court system isn't really very fluid. It's there isn't contextual guidance, there isn't any um help outside of the help you can get from a lawyer. Judges put a lot of effort into helping people. Uh, when they when they see that one of the parties isn't comfortable, very often the judge will try to help them out. Um, but they're limited in how much they can help because they have to remain impartial and independent. So that is also an obstacle for some people. Then you have a lag that is inherent to the legal system. Um, the legal system cannot preempt innovation. The legal system has to adapt to what to how the society evolves. So if we see that the market is trending towards a certain direction, that doesn't mean that we can immediately engage the legal process, come up with new regulations, revamp the courts. We can't do that. We have to make sure that before we start changing the legal system, the trend that we're observing and which we want to adapt for is here to stay. We have to make sure that the changes we're adapting to are permanent to a certain extent, of course. Um and finally, the language, the legal language is not easily accessible. We've done a lot of progress, uh, we've made a lot of progress in helping people navigate this. There's a lot of help online available, the chatbots are multiplying, um, but it's still a formalized process. And using the wrong vocabulary can sometimes have a detrimental effect on the rights you have. So I would say those four elements are the main obstacles to people feeling comfortable using court systems.
SPEAKER_06:Yeah, and in language, um, that is so much important to make sure that first it's a balance. You want to make sure that people understand what you're saying or the stand their the the full stand of the legal process, but you don't want to use a language that is deceiving as well, because as we know as lawyers, synonyms in common language can be totally different things in under the law. So it's it's it's a it's a tricky balance to find.
SPEAKER_02:It's a very tricky balance, and sometimes you don't have the choice. If you don't use the right terminology, then you're not enforcing the right rights or you're not you're you're not defending yourself effectively, you're not presenting your case in the most persuasive manner to a trained legal-minded judge.
SPEAKER_05:Yeah.
SPEAKER_02:Um, and likewise, if you have another, uh if you if your opponent um has a lawyer that may want to act a bit aggressively, um, which I want to mention is completely okay because they have to defend the interests of their client. That is completely fine, but they also that also means they have to exploit it when you make a mistake.
SPEAKER_03:Yeah, of course.
SPEAKER_02:Because that's to the benefit of their client, and we want lawyers fighting for us. So those are hurdles.
SPEAKER_06:Yeah. It's true, it's true. It's it could be meat for them to bite on.
SPEAKER_02:Don't give meat to the sharks.
SPEAKER_06:Okay, it's true. So you have um helped build uh e-arbitration platform and have worked with AI power tools. What's the most promising way AI can simplify access to display resolution for people and small businesses?
SPEAKER_02:There are several areas. Um, first of all, there's the kind of dispute, and then there's the general process. So maybe I'll focus for now on the general legal process. The first thing is user centricity. We can build the dispute set system around the needs of the users because now we have really advanced uh design methods, we can really create effective interfaces, we can make things easy to use for even for people who are not accustomed to using new technologies. And this is something we should leverage. We this is typically the field of legal design. We know that by building effective designs uh for legal work, we increase engagement from our uh clients. We know that C suite uh decision makers will more easily engage with the legal documents if it's been formatted using the UI design principles. UX, sorry, design principles. Um we know how to simplify terminology so that the person knows what expect what's expected of them. Uh we know how to design ergonomic workflows um so that if you if you're wondering what you need to do, then you can understand from the interface that you have in front of you how you should do that. That is really critical. Uh second, we need to use data. We've been we've been creating a really large amount of court data. And I've been advocating for two, three, I can't remember, a couple of years, three years at least, uh, that we should be leveraging court data to inform decision making. It's not saying we're doing copy-pasting, it's saying these are the odds, or this is how a dispute like the one you're experiencing gets resolved in practice. You can leverage that. You can leverage that to improve transactions, settlements, you can leverage that so that you ensure that there's an equal application of the law for everyone. You can detect bias, you can detect perhaps personal uh prejudice uh from uh from judges or from from um from case law in general. Um it's really important to leverage all this data that we're accumulating so that we can help people better resolve their disputes. It's not usable the same way for every kind of dispute, but every kind of dispute can benefit from those, from that data in one way or another. And then we have to think about the outcome. Probably we need to make dispute resolution faster.
SPEAKER_05:Okay.
SPEAKER_02:One of the reasons why is because if people believe they can have a fair decision in a short amount of time, they're willing to take more risks. That is generally a driver of economic stimulus.
SPEAKER_05:Okay.
SPEAKER_02:You're willing to take more risks if the risk you're facing is reduced because you you have trust in a system that it's going to be able to adjudicate the dispute quickly and fairly. And you'll put your investments will be protected, your economic activities protected. This is something that really helps. Um, and then you also want to be able to address claims faster because sometimes people are in a very tough spot. And if you're not able to get them out of that tight spot fast enough, they may encounter issues that are really difficult to get out of later. So speed, outcome, quality, all of that is a very aspects that we can work on with the new tools available and leveraging the data that we're accumulating.
SPEAKER_06:Yeah, because trust um in the system uh makes everything else a bit easier or a bit uh you it's something else uh less to be worried about. Um and then the the speed, the efficiency, and the being able to navigate without feeling lost is is is quite uh is is is is quite an endeavor, and it's something that is not very yet attainable right now for most people.
SPEAKER_02:It's getting better because I mean you have some startups building chatbots that can help navigate. Okay. We have some lawyers that are rethinking how they're offering legal services, um we have some judges that have that are more sensitive to the struggles that people uh can be facing. But we now have the tools to really take a step forward. But as I said, there's lag between what the courts can do and how fast they can do it. Because in any case, if you want to change the court system to a certain extent, this becomes a political discussion.
SPEAKER_06:You need to change the law.
SPEAKER_02:Sometimes yes, sometimes no. But in any case, it becomes a political discussion, and you need to make sure that we reach a consensus so that the solution is acceptable and we don't get challenged every every X years by somebody who wants to, well, who feels that they haven't been heard and that they have valid claims.
SPEAKER_06:So this is a good segue to the next question. Um, there's often tension between making justice faster and making it further. How do we use technology to balance affordability and death, especially when public trust is fragile?
SPEAKER_02:So public trust, I have to say we're quite privileged in Switzerland because of the consensus the consensus building system. Everybody gets to be heard. At least a wide majority of the population gets to be heard. And the fact that it's a small country means we know each other quite well. So that really helps to tone down the uh potential uh uh rifts that can build over time. But in any case, um the first thing is AI acts as a multiplier in in many cases. So the studies that have looked into how AI affects legal work, as far as I'm aware, there's only one study that was really done outside of the OpenAI uh Microsoft environment. So one at least presumably transparent independent study that showed that the main impact is speeding up the legal work. The study dates back to 2023. So those are tools that were widely less capable than what we have now. But those tools still in 2023 allowed for a gain in efficiency of almost 23%. So you can be at least 23% faster. And that's not nothing, that's almost a quarter. Um then it allows you to highlight patterns. AI is really good with pattern recognition. So if you all of a sudden want to know whether your court system is really truly fair, um, if it's somehow affected by prejudice in certain places, in other cases, you you can audit your case law using AI to try and pinpoint where the system isn't operating as it should. That's a really beneficial way to use it. Um obviously you always have to keep a human in the loop. You cannot just let the reins go.
SPEAKER_03:Okay.
SPEAKER_02:But you can enhance the capability of the human actors of this of the court system by giving them access to more information, to more insights. Um, for example, if a lawyer needs to answer their clients to explain, well, what are the odds of winning or what are the odds of losing? Big question that they always ask. Exactly, because the clients want to make data-informed decisions. Of course. And if if lawyers just told clients, we've had 10 cases like yours, we won seven of them. On average, it cost us between 8 and 12k, or it took between seven and 10 months. You are giving data. Yeah, and it's much better for the client than simply saying your chances are good. Um and the price and time, it depends on the judge.
SPEAKER_03:Yeah.
SPEAKER_02:You need to give the clients data because that allows them to plan.
SPEAKER_06:Yeah, of course.
SPEAKER_02:And that allows them to measure risk. So if you are actually using the leveraging the data of the court, then you're more likely to be able to convince your client whether it's a good idea to go forward or not. If the client decides not to move forward, which they can do today if you say that their odds are not great, which you sometimes have to tell them as a lawyer, yeah. Well, maybe they'll think, oh, but was it not great 49% or not great 9%?
SPEAKER_05:Okay.
SPEAKER_02:Maybe it's 49%, they would have said, No, I still want to take a shot. They still want to take a shot.
SPEAKER_05:Okay.
SPEAKER_02:Because people have a really hard time working with a spectrum.
SPEAKER_05:Yeah.
SPEAKER_02:They have it's very easy to work with zero and a hundred, but working with 23.4, that's really complicated.
SPEAKER_05:Yeah.
SPEAKER_02:So if you tell most people you the risk is low, they'll just assume it's zero. But if you can put a number on it, then it allows to put the the the answer into perspective. So that really helps. And likewise, if you inform a client that their claim is unlikely to succeed, they might decide to go see a lawyer because they think the lawyer will increase their chances of success. Of course. Because you can present the case differently. The client doesn't know what's the best way to present a case. They don't know the judges, they don't have the persuasion that a lawyer may hold over a judge. Some lawyers have an incor incredible reputation that they can leverage to the client's benefits. So you might want to take that risk, and you might want to decide to actively consult a lawyer because you think that's really going to make a difference. So there's it's it's very difficult to see how it impacts in practice, but there are a lot of moving parts in this.
SPEAKER_06:So it's it's um it's a numbers game. That would be the overall idea. Uh, but also is the the having all the information at hand. Because as you said uh before, some people are more risk-averse, some others are more risk risk takers. So you have someone say, Well, you only have 2% chance of winning. Say, Yeah, but I have a 2%. They will see it like that and say, We're gonna do it because I I believe it's the right uh it's the right uh course for me or for my company. And then you will have people, oh, you have 80% chance of winning. Saying, Oh, so there's 20% that we don't maybe we don't pursue it. So it's uh and having the privilege of information, I believe, is is the most important part here to make sure that you have all the tools and then you decide if you use them, use them or not.
SPEAKER_02:Yeah, definitely.
SPEAKER_06:Perfect. I I like that. Hopefully, it will be something near in the future that we can assess and say, oh, how many cases about this and this uh has been brought to the court.
SPEAKER_02:Yeah, I think it's already happening in some jurisdictions. Um, I mean it's it's the logical next step to digitalizing the court system. And we know that some states have already done that in parts, some institutions are doing that. We know that it's the next step. It's just a logical next step to take because you're building uh case data. You're building data. So then the next question is what do I do with it? How do I use it? And most judges in the world have a case law, uh a backlog of case law, that um a backlog of cases that is really difficult to handle.
SPEAKER_06:Yeah.
SPEAKER_02:The most justice systems are understaffed.
SPEAKER_06:Yes.
SPEAKER_02:So the next logical step is can I leverage that data to lower the work, the backlog, make sure that people get an answer quicker. That's really that's really a very natural way to move forward.
SPEAKER_06:To make to make process that can be more efficient, to make them more efficient.
SPEAKER_02:Exactly.
SPEAKER_06:Making sure that for example, uh I'm at midwork, that can be uh a process that the documentation, the evidence, all that can be done faster and more efficient ways.
SPEAKER_02:Yeah, exactly.
SPEAKER_06:So that that can help to make it faster, and also for for the same part of the um understanding the evidence, all this elements of the evidence is also can be done by AI, of course, with a human eye, surveilling, making sure that everything is accurate, but it will make it faster and easier to understand.
SPEAKER_01:Yeah, definitely.
SPEAKER_06:Then we're going back to design. So people talk about legal empowerment, but the interfaces are still designed for lawyers. But we're not that bad, but we're not everyone's cup of tea either. What role should design and user experience and language simplification play in building AI system for legal access?
SPEAKER_02:We've seen very successful cases. Um I was mentioning before how legal design had helped um lawyers communicate with decision makers in companies, how in-house counsels sometimes also use legal design to make their advice more intelligible. Intelligible. Um, but we we we observe that it results in higher engagement. We observe that it smoothens the dialogue. Um, and we've seen it already happen in the medical industry. We've seen uh doctors being actively trained to avoid the medical lingo when talking to patients.
SPEAKER_03:Wow.
SPEAKER_02:Because the patients need to understand what's what's affecting them as clearly as possible.
SPEAKER_03:Of course.
SPEAKER_02:And we've been doing that as well in the legal industry. A lot of lawyers have been sensitized to um to the need the communication challenge with the public. Um and we've seen a lot of uh uh projects by different lawyers uh across the world to try and bridge that communication gap. So that's really important. User experience specifically is a very interesting tool to use. User experience is basically trying to bridge the communication between a human and a machine. This is not just some drawing on a board. You can have a PhD in user experience design. You cannot have a PhD in drawing.
SPEAKER_03:Okay.
SPEAKER_02:That's the level of difference. What you're using is psychocognitive um psychology.
SPEAKER_05:Oh.
SPEAKER_02:Uh sorry, uh cognitive psychology. You're using cognitive psychology to understand how a human is going to interact with an interface to make sure that they can accomplish the task that they have in mind when they're performing an action.
SPEAKER_05:Okay.
SPEAKER_02:That's why when you have a car, the brakes are not in the trunk, the brakes are under your foot because that's where you have the most power when you need to stop a car. It's not in your fingers. If you can stomp on the brakes, that's the most effective way to stop a car that doesn't have ABS or any kind of AI assistance in it. And that's how we started it.
SPEAKER_03:Yeah.
SPEAKER_02:So that's why this tool is so powerful. I mean, you you almost don't have any software nowadays in large corporations that comes out without a really intense UX review.
SPEAKER_06:Of course.
SPEAKER_02:Um, and that's how Google, I mean, search engines went from dictionaries.
SPEAKER_07:A thousand things.
SPEAKER_02:It was really really overloaded pages.
SPEAKER_07:Yeah, very populated.
SPEAKER_02:And now we just have a search bar.
SPEAKER_07:Yeah.
SPEAKER_02:And the search bar has been improving over time, but it's one of those very few cases where incremental improvements didn't result in losing people or people complaining that we're adding useless features, etc. It's just getting better and better. Yes. So applying UX design principles to legal to the legal system in general can definitely help make it more accessible. And UX does include simple language. Um, if you're if you're thinking about buttons, well, if you have a button called file a claim, it's not as easy to comprehend. It requires more brain calories than if you just have submit or I'm done.
SPEAKER_03:Yeah.
SPEAKER_02:So you try to figure out what's the best way for the user to understand what they're doing and what the impact of their actions are going to be.
SPEAKER_06:And symbolism is very important as well. Making sure that you have the right symbols that convey the right message.
SPEAKER_02:It's actually not that natural at all to read. We're not reading creatures. Um instinctively, we respond better to action, to stimuli, to visuals, because we were designed that way.
SPEAKER_07:Yes.
SPEAKER_02:Um, reading is a very uncommon characteristic in the natural environment. Yeah, exactly. It's just what, 5,000 years old reading? No, maybe it does, but at least it's Babylonian, I think. But anyway.
SPEAKER_06:Um a few less. We were visual animals first.
SPEAKER_02:Cuneiform. But it's definitely not something that we respond as well to as clear visual indications that we can comprehend because they they tell us what needs to be done.
SPEAKER_03:Yeah.
SPEAKER_02:Um, a very simple example is a dog. If a dog comes and scratches your leg and then it sh and then it walks to the door, most people understand better what that means than if the dog were able to write, uh, I need to go out. The reason is because you don't know why it wants to go out. But by grappling your uh by scratching your your leg like that, yeah, there is an in there is um uh an inferred um notion that it's important and then it's urgent.
SPEAKER_05:Yeah.
SPEAKER_02:And you also know doc behavior. So you're just gonna try to figure out what they want. And if they just walk to the door or to the to the water dish or to the food dish or whatever, you're just gonna understand what they want better. It's more, it's easier to understand if you can use the correct visuals.
SPEAKER_06:Yeah.
SPEAKER_02:But that again, it's also culture sensitive.
SPEAKER_06:Oh, interesting. It's interesting. I it reminds me of I have two kids. And you know, when they are babies, when non-verbal they cannot speak, so you have to guess what they want. But then you start to realize that there are different cries and there are different ways that they complain about things and that there are nuances to that. And then you start understanding, even they're they're not talking yet, but they're telling you what they need, and you you learn as well because you you you by error and and uh you start learning how to better understand what they when they what they're trying to tell you. And this is something that's um uh visually um it it comes more natural. It comes more natural when they point at things or they tell you what they for my my youngest son, he points at the bottle when he wants uh milk, for example. Uh but I uh I was saying in my oldest, when she said milk, milk, me, I say, Boy, do you really want milk? And because she's three, it's like do you do you want something else? So I negotiate with her, so it's different. But I think my son's way is more efficient because he only takes me and I go immediately go and give him the milk.
SPEAKER_02:Yeah, yeah, and you can see it also in LinkedIn posts nowadays that pointing thing you mentioned. Yeah if you just write link in the comment below, it's not as if. effective as if you put a finger pointing down. I can't tell you what the exact data points are, but it is more effective.
SPEAKER_06:Yeah. And also the carousels are on LinkedIn themselves. Like the carousels, I find them way I have more engagement whenever I post something with visuals, carousels or post with any image than when I put only text or when I put only a link. So it's there's less likely people are less likely to even um like wait oh I want to see this. So this is it's human nature we we are visual the carousel is a great example.
SPEAKER_02:And you can also see it in the ad requirements on the large advertising platforms. They generally prohibit you from putting too much text because they know there's not enough engagement.
SPEAKER_06:Yeah people are going to get bored. I'm going to pass it so talking about uh technology so and tools many legal text uh tools focus on helping lawyers and the big firms there's nothing wrong with that we know there's a lot of money to be made in that realm um but also AI can be more efficient for the judiciary have we been talking already so make it more efficient financially sustainable for the people using the judicial system so what does profitable justice mean in this context so we were talking about before of the problem with underfunded justice I believe um AI is a good way to generate income for courts by leveraging court data.
SPEAKER_02:There's a lot of ways to do it you could charge a subscription for accessing simulators to try and assess what your what your chances are you could but what I think would be most effective is offering an online small claims uh settlement service how that would work it would basically be a person submitting the details of their case to the on the platform having those case details analyzed by an AI matched against a database of cases and return a settlement offer based on the existing data. It's like for a case like yours this is how the parties usually settle.
SPEAKER_03:Okay.
SPEAKER_02:That is something that you can charge for a very reasonable price compared to the price of a more extensive uh case and you can reinject that money into the court system.
SPEAKER_03:Okay.
SPEAKER_02:Of course you always have court oversight of course you always have the ab the the uh the the the the possibility to submit your case to a human judge but you're able to offer an online service that that is in that is targeting the per the the the the people that don't use legal services. At the Swiss Eagle Tech we had discussed once with the Ministry of Justice of South Korea and they had been able to measure how many uh what was the percentage of the Korean population South Korean population that didn't use the legal system. And it's a very large amount compared to the people who use the system. Okay. I don't know what the data is for Switzerland. I don't believe it exists but if we realize that a significant share of the population doesn't use or doesn't engage with the court system when they have an issue these people are it in effect deprived of justice. Yeah and there are many ways that we can think about a system where these people can be included um they can generate income without overloading the court system even more and allow the court system to offer a solution to everyone. That's really the objective we're looking for the really the key criteria is can we do it at a price where the maintenance fees do not exceed the income.
SPEAKER_03:Okay.
SPEAKER_02:We have to find a way to make the service affordable but it has to generate enough income to sustain itself ideally a profit as well. And this is something you can do on a national level so that you can leverage the largest possible population so that you can generate as much income as possible and so that you can reinject it in the legal system itself.
SPEAKER_06:Yeah because small claims is has proven to be a successful uh process all around the world and it really helps especially small businesses um independent people or freelancers who just want to get on with their work and and keep trying to make a living um or or growing in their own um businesses or or their own art so and and this provides them with uh access to justice so it would be interesting to find that tool to be it uh online more accessible more easier to to understand and often in this kind of course you don't need a lawyer to represent yourself you just need uh someone to guide you a little bit and you can also do it yourself is it's if it's not a very complex case I think we may be seeing so specifically to the point you mentioned you don't need a lawyer.
SPEAKER_02:I'm not even sure that you don't need a lawyer because again people there is there is this there is this notion that you don't understand the lingo.
SPEAKER_03:Okay.
SPEAKER_02:So it's not clear that people would not use a lawyer to make sure that they're presenting their case in the best way possible. Because obviously an AI would respond better to the legal lingo because that's how the case law is built. Of course so you would still potentially need a lawyer but then the question becomes what kind of service are you getting from a lawyer? And I I could very well see a new type of service offered by lawyers um which isn't the traditional activity that we think about. It could be simply okay I think my interns need to be trained. We have a very big challenge in training the newer generations um and the the objective for trainee lawyers is to see as much of the practice as possible is to see as many cases as possible so that they can really understand the profession and and develop their skill sets. Well this could be something for trainee lawyers that can be reviewed by their uh supervisors. Um and then the service might be a short review to see if you're doing it the right way.
SPEAKER_03:Yeah.
SPEAKER_02:And of course the pricing would be adapted to the service it makes no sense to charge$2,000 for that. But you could also consider different kinds of pricings. You could also be a nomad lawyer with very small expen um very low level of expenses you could be living in a in a jurisdiction where the cost of living is lower. And I know that some lawyers that we have in in Switzerland just work like that. They they expatriate and they work from a distance this is a perfect system for them. Likewise it's a good system for training young lawyers. It's a good system for filling in the hours where you're not actually working.
SPEAKER_03:Okay.
SPEAKER_02:Uh it's a good way to develop your network by word of mouth. So if you help somebody with a small case they might think of you as well when they have a divorce or something that requires a bit more time. You might see more engagement with lawyers in general you might reduce the barrier to to consulting with a lawyer. Okay. So that it can it can work out in very different ways. But if you're making the legal system more accessible then the market organizes around that. And it's very interesting to see how the market can respond to that. And generally speaking it's the market organizes in a way that makes it profitable for many.
SPEAKER_06:So always need a lawyer or always you always need a lawyer of course you always need a lawyer but on your own terms.
SPEAKER_02:On your own terms yeah you you you have the right to choose your lawyer that is a that is a fundamental guarantee of the legal process. Yeah so I think lawyers stand to gain if we diversify our offer to more people it doesn't make economic sense to all fight for the same clients because we're just dragging the prices down. Yeah it makes more sense if we specialize in a certain type of and we already specialize in anyway in certain displays but it also makes sense to specialize to a certain client base. And if you're charging eight hours a day at 100 per hour you're making as much as if you're not if you're charging four hours a day at 200. And then it also depends on how much income you need um I I know a couple lawyers who enjoy the high complexity of cases of mergers and acquisitions uh really tough cases but there are also lawyers who really help who who really like helping people who need legal help. Yeah and if the system also allows them to operate effectively then it's a win-win.
SPEAKER_06:Yeah because they will be more inclined to take more cases like that I definitely don't want to take any inheritance case I don't like that for me it's it's LAN anything regarding LAN or oh good lord yeah we all have our babies terrible terrible terrible it's well um nothing wrong with it but no it's just you like it's just very very boring I like tech I like contracts like IP I like IP it's technology creativity I like dynamic yeah yeah I completely understand that I think for everyone will find their passion and if we have tools to improve that it's it's better for everyone yeah definitely so should we imagine courts more like services always on digital first or do we risk losing the human core of justice when we go fully online it has to be somewhere in the middle okay it has to because the court system is working but it's not in line with today's requirements today the trends are being set by companies like the the those American giants um possibly tomorrow they'll be set by Chinese giant giants um but the matter is the fact of the matter is our ways of consumption have evolved and a lot of services are being democratized.
SPEAKER_02:Yeah they're get becoming more accessible to more and more people who would have thought that your average Joe could invest and trade that I mean that was unthinkable 10 years ago and then online small trading platforms started appearing and being successful. You had crypt crypto platforms that developed you have new technology coming all the time for to target the the the the general public to make their offerings available to the the the the the widest group of people possible that means that inevitably we're going to have to rethink how rethink how the law services legal services are offered can we make it a lot better yes we do have a challenge to make it easier faster cost effective yes but we cannot get rid of the human and it's not going to happen for a very long time because AIs just don't have the capacity to replace humans. And right now let me say that we're getting AI at a really cheap price compared to what it actually represents. Yeah and I mean when Uber came out it I mean I saw that on LinkedIn the other day it was a very very relevant comment but when Uber came out the the the the the rides were much cheaper than they are now yeah I'm pretty sure it's going to be the same for AI. They're gonna hook us and then well yeah at least we might get used to AI being very cheap and we might not be ready for a price hike because I mean open AI is losing bucket loads of money. It's like an unending fountain of money pouring in but it's also a really holy bathtub.
SPEAKER_05:Okay.
SPEAKER_02:So at least contrary to the Uber uh description you have so many companies doing AI now that the scenario is not comparable.
SPEAKER_06:Okay.
SPEAKER_02:Um but we still get AI very cheap at the moment. But it will get cheaper it might or it might not because a lot of companies are not making enough money to justify the investments. At the moment we're in a gigantic race at at world level to who's gonna set the standard so who's gonna say the standard who's gonna get artificial general intelligence first. This is basically to put it mildly that's basically the almost complete automation of human labor. But it also means the biggest geostrategic weapon that we've ever had but an AGI can achieve roughly the same results without killing anyone without destroying anything I don't know if that's better or worse. A hundred times worse at least good lord the reason why is you can coordinate attacks on infrastructure you can coordinate disinformation campaigns you can you can completely paralyze the financial system you can absolutely break all the infrastructure everything can be paralyzed on a systematized attack. Okay and it's unbeatable by a human because of the computational speed of an AI. Needed yeah so that's why that thing is actually in the crosshairs of all the major powers today. And they are very clearly saying that that's what they want to achieve human-like intelligence.
SPEAKER_05:Okay.
SPEAKER_02:And the way we get there is obviously building the most advanced AIs but by coupling it with quantum computing quantum computing is turbocharging your average processor exponentially and I'm not making this up Microsoft's announcement um earlier this year I think in March with their Majorana processor they want to go up to one million qubits. So qubit is basically the quantum computing version of bits um if you have one million qubits you have a ginormous capacity to process to calculate you have unbelievably powerful processors. It's not comprehensible by the human mind. We're talking about numbers that go beyond the billion billion billion times faster. It's not possible for the human brain to comprehend those levels of magnitude. So that's what we're dealing with. So we're quite close to the general intelligence we're quite far actually because quantum computers nowadays go I think up to a thousand qubits but they're very unstable because to get that thing to operate I'm sorry if I'm getting too no no it's it's okay I it's interesting because on the intangible in Spanish part we we talk about quantum computers so it's good that we have that also on this side so if you have um a quant to to get a quantum computer to work you have to be able to control subatomic particles meaning particles that are smaller than an atom if you want to be able to control that you need really very finely tweaked um instruments and it's very difficult to control a large number of them to get them to operate without error um in a coordinated way. Let's put it like that. So we can do we can we I mean we're we're having breakthroughs every now almost every three months at the moment but we can we can already do that but it's still out of reach for now with our level of technology. And AIs are having some significant challenges of their own to overcome um chat GPT regularly forgets the instructions because of the size of its context window. So now we've we have much bigger context windows sure but if we want to process that level of data we're nowhere near what we need. Will we make breakthroughs very quickly yes we've been making breakthroughs at an alarming rate um but we're not there yet okay and we've seen I mean we've seen the tension there is on the market a very good example of that tension is when Nvidia lost six hundred billion dollars in a single day when Deep Seek was announced because people thought oh actually they could do it for a hundredth of the price so it's it's a very and on top of that we we have this trade war going on in the background we have political tensions rising because of autocrats everywhere autocrats are really prone to go into conflict they like conflict um we very uh we very rarely see an autocrat having a cup of tea and chilling by the fireplace okay um so those profiles generally tend to want to project force and that means conflict so we don't really understand how the parts are going to play out we could have we could be looking at a couple years before we have something really dangerous or we could be looking at a couple decades.
SPEAKER_06:Okay.
SPEAKER_02:So we will be governed by superintelligence by the end of the century oh I wouldn't make any prediction for the end of the I I hesitate to make a prediction in five years.
SPEAKER_06:Oh wow okay um but I would actually like to be governed by a superintelligence it's just that statistically speaking a superintelligence is very likely to kill us all yes of course we are not efficient we are not we are hindered to the environment we are not uh all good in nature seeing it yeah we don't know if an AI would see it the same way but the AI is trained by us it's gonna learn that what we think of ourselves so there you go.
SPEAKER_02:But not necessarily because it's uh it's very difficult to project how the advancements of the technology will impact its perception of its environment. Okay. It could also consider that you know do you care about ants when you're walking in the street not really you only care about ants when they start invading your kitchen it's true when they're right in your face. We don't really know but we have a tendency to personify the the AIs with our own selves. We have a tendency to project how we would react on the AI.
SPEAKER_06:Well that's true.
SPEAKER_02:And we would definitely see the AI as a risk it's just that we're building it so of course we we have a bias in favor of it because you don't buy something you find you you don't build something you find stupid.
SPEAKER_06:Exactly I I like what you say that it's a reflection of us as well. It is because it's often like that many times even with human interactions when people get sometimes offended it's not what about what you say sometimes it's what what they think you think they think so it becomes a whole thing or they're reflecting on or they already have an internal dialogue and then you you somehow touch that and then that triggers something in that person but you not necessarily mean to say it like that or you were talking about something completely different. Yes exactly okay I I haven't thought about it like way but yeah it it it's um it's a reflection of us of what we think of ourselves as humans but maybe it will be merciful with me I don't know because I often I often fight with mine.
SPEAKER_02:I'm always fighting with uh especially with the open AI one I'm always fighting saying like no this is not like this is not like that you're having uh conversations with uh an inanimate object yes I am I am yes uh it's it's a new it's a step up of talking to myself at least you're good company you know I mean at least you understand what you want most of the time exactly kind of it's complex amazingly annoying discussions with AIs every now and then it's amazingly infuriating yeah because then at some point it tries to patronize you oh my god yes so annoying and and often also when I say but this is offensive it's like ah I didn't see it and I was like come on you really didn't see it you're just dodging exactly typically so um the last question of this segment um if you were asked to advise a government on implementing AI in courts what would be your top three principles to ensure inclusion efficiency and long-term viability that's a really tough one I think the number one aspect is distributed authority you never want to have only one person making the decisions you need to have some consultative body that has enough independence to be able to operate so we know how that works in human societies but we still have to try um we need to ensure that we have accountability for decisions that are made we cannot have people just being immune for whatever reason in the world because they don't like accountability whoever does like accountability anyway but we need it okay um we we absolutely need accountability and we need to be able to enforce it um so and and the final thing we need to maximize inclusivity and um in in the design of the system so you need to have different profiles to understand the full depth of the system um you you need to you need to have lawyers you need to have judges you need to have politicians you need to have technicians you need to have architects you need to have designers you have you need to have psychologists anything that can add a layer of understanding to the um to the system is going to be beneficial then you also need to ensure that your data is considered a top priority strategic asset you can't afford to let that data escape you cannot afford to depend on one specific provider because that gives them too much leverage uh you cannot afford to depend on one infrastructure for backup you need to have multiple backups you need to make sure that the whole security aspect is considered the privacy aspects are considered the ethical aspects are considered all of that means you need to tackle the problem in a really responsible way ensure accountability and ensure that the accountability is enforced on a personal level you shouldn't be able to dodge accountability because you're behind a company okay okay so fairness I mean you have to take it seriously using that core data to actually implement um a a new way of working requires a really high level of security of ethics of compliance of course um and it's better to be overly cautious than perhaps a bit too complacent.
SPEAKER_06:Perfect I I I like that uh part of of uh making yourself responsible and and making yourself accountable for the decisions that you make or you let make according to the system you have created so now we're gonna go to the fun fast part of the of the episode I'm gonna tell you two options and you will have to pick one without thinking too much we have uh let's start let's think fast pick one AI powered small claims court or human judge with nine months delay you can't just add the nine months delay like that you need a small claims court but you also need the judge with nine months delay okay you need both you need both okay but if you have to pick one you can't I mean I I would be doing a disservice to my message true true but you can make the judge more effective and you can answer in six months or five months three months that's what you're aiming for.
SPEAKER_02:To push it down to but to yeah to push it down and make sure that the the the judge is equipped with the tools they need to be able to reach that level of performance without over you know without burning out of course yes of course because judges are humans they are really good humans too yes of course they're but they they went into the profession probably because they they believe in justice they do and I mean for every time I discuss with judges um I think they all welcome an uh an opportunity to find the best possible outcome for a case okay uh something that that is consistent with the law and that is fair.
SPEAKER_06:That's beautiful that's beautiful it's it's something that it keeps hope and the hope alive as well.
SPEAKER_02:You know most most public servants they they really have their our best interests at work.
SPEAKER_06:I love that so a chatbot that resolve 80% of the issues or a human lawyer most can afford most can afford that most people cannot afford again you need both yeah because 80% of the cases might not go to a lawyer at all.
SPEAKER_02:So in effect they cannot afford that lawyer but those 20% of cases you can afford the lawyer and you need it. So it's it's it's it's a horribly difficult you're very swift in your answers. It's really true right I mean if I try to look at the data it's probably true. Yeah um I mean as much as I like a chatbot it's not equipped to deal with everything um and an unotated I mean an unaffordable lawyer most of the time they they deal with really big uh legal cases so they are by now by nature unaffordable to most people but there are other types of lawyers as well available there's pro bono there's other types yeah but those are not unaffordable true I'm going out of my own questions okay so next question radical simplification of legal language or keeping precision even if you understand it yeah you would need to keep the precision but you would be able to use AI tools to simplify it. Oh but you need the precision. Okay so there's a twist to it but but you but that's what we're doing now we're using AI to simplify it so that it's accessible. But if you if you use the analogy with the medical world again it doesn't help the medical profession if we stop using the the technical vocabulary um we need to be able to use um the the the the proper name of a bacteria um but at the same time we need to be able to explain to the client so you need the precision but then you need also the the the translator the the language for for everyone but without losing the precision.
SPEAKER_06:Exactly again both men oh god you're it's this has been very very swift answers very neutral it's important okay next question I Hope you pick one public access to digital legal tools or protecting them behind paywalls to fund the courts.
SPEAKER_02:You need both. You need to charge us a fee for the service. Um, because you need to fund the system. Otherwise, we know what it is, otherwise, we just don't have a system at all. It just doesn't operate. Okay. But the question is, where do you strike the balance? And I don't have the answer to that. It needs to be analyzed from an economic standpoint, from a financial standpoint. It's really not something I can say one or the other. Okay. Okay. Of course, I prefer free access to courts for everyone. I would prefer that, but it's not realistic.
SPEAKER_06:The courts need to be paid. Yes.
SPEAKER_01:And the resources and the infrastructure and everybody.
SPEAKER_06:Yeah, of course. It's a system that needs to be sustained. Oh, God. Okay, let's see. This next one is the good one. Open source tools for justice or proprietary platforms for control.
SPEAKER_02:Oh, open source.
SPEAKER_06:Okay. Ah, we have one.
SPEAKER_02:Oh, you can't you cannot have proprietary technology because you're just locked out.
SPEAKER_03:Okay.
SPEAKER_02:And you don't have transparency, you don't have accountability, you cannot have a community-driven improvement. Open source for that kind of software, is my view, is much better.
SPEAKER_06:Perfect. See, he picked one. Okay, then we have free legal access for all, even if it means more false claims or tighter filters that exclude some real needs.
SPEAKER_02:I like the first option. Okay. Because we already have mechanisms in place to deal with frivolous claims. So the procedure code, the the the code of silver procedure or the criminal procedure, there is always something to deal with frivolous claims. So I'm perfectly happy with free access.
SPEAKER_06:Okay, free access. Yay, two in a row. This is great. Redesign the courts online for online access or redesign society for fewer people need to use them.
SPEAKER_01:You need both.
SPEAKER_02:We're back to neutrality. It's not neutrality, it's you need both. Okay. Um, although it's much easier to simplify the court system than to tailor society to not need so much assistance, um, especially because those kind of adjustments occur over an extremely long period of time. It takes a really long time to educate a population on a societal level. So, in the meantime, you need to have easily accessible systems, but it's not an excuse to do neither. You really have to work on both because every country stands to gain when the population becomes more proficient at something.
SPEAKER_06:Yeah. Make sense? You make your case. Empower users with AI to understand the law themselves, or invest in more human help to walk them through it.
SPEAKER_02:I would go for uh empowering users with AI with the notion that you will have people available to help. Yes, you will have them because it's a market. Okay. And and some people have specific needs. You cannot address all the needs with a sp with just one tool. So at least that for now you can't. Um But yeah, you would probably need help on the side as well in any case.
SPEAKER_03:Okay.
SPEAKER_02:And I think it's also a healthy system because it promotes best practice development, it promotes um the development of good tools in general. So yeah, you have advisors, you have users. I think those two work really well together to improve a system in general.
SPEAKER_06:Okay. And finally, courts that adapt to people's lives or people who must adapt to the courts.
SPEAKER_02:Yeah, it has to be the courts that adapt to people's lives, but there is a lag. There is a lag. So in effect, it would be good if the courts were always aligned with what the people need, but it's impossible.
SPEAKER_05:Yeah.
SPEAKER_02:So the people have to be able to use the court system. Um, and the court system can follow with the lag that is inherent to how it operates and how any institution uh is going to operate, uh, especially in a country like Switzerland with a lot of diversity of language, of um, of culture, um you're you're you're you're going to have this lag anyway. So the the primary objective is to make sure the people feel comfortable using the court system, but over a longer period of time to get the court system to follow up and make itself more accessible. So you have uh one party doing a lot of the road and the other one doing a bit of it.
SPEAKER_06:Okay, okay. Yeah, because the the law is meant to serve the people.
SPEAKER_07:Yeah.
SPEAKER_06:And the people change, the culture change, society changes, and we evolve, we have different needs, different understanding of what we need.
unknown:Yeah.
SPEAKER_06:So it's important to follow closely. Yeah, yeah, you have to. Okay, now the game part. Take the little paddle there. So on one side you have um on one side you have true, and another side you have futuristic. So I'm gonna give you a statement and then you tell me if you think it's true, meaning that it's happening right now or it's about to happen. And the other side, futuristic, if you believe that it's something that is far in the future or is never going to happen, or is false. Okay, first question: Are you ready? Citizens will file and resolve simple claims from their smartphones without a lawyer.
SPEAKER_02:That's true. It's already happening.
SPEAKER_06:Okay. National courts will offer express automated resolutions for minor disputes.
SPEAKER_02:Yep, I think it's already happening as well.
SPEAKER_06:Okay. Multilingual AI assistance will replace legal aid.
SPEAKER_02:Yeah, futuristic.
SPEAKER_06:Legal information will be distributed via voice assistance to people without literacy access.
SPEAKER_02:Yeah, that's already happening, yes.
SPEAKER_06:People will trust virtual legal advisors more than traditional ones because they actually understand them.
SPEAKER_02:Better.
SPEAKER_01:Ah okay. True.
SPEAKER_02:The problem is, I I think it it's worth it's worth saying it. AIs have an amazing ability to understand a person's emotional state.
SPEAKER_03:Okay.
SPEAKER_02:Um, and it overperforms humans, it outperforms humans. So being understood, the AI might very well give you the feeling that it understands you better.
SPEAKER_03:Oh wow.
SPEAKER_02:So this is already a reality today. Um will the AI trust be trusted more in a court case? I think it really depends on what we're talking about. If you are um the victim of harassments on your and on your employment um uh at the office, you need a human presence near you in some cases to attend conciliation hearings, okay, uh, to attend court hearings. The AI is not gonna give you that reassuring presence and speak up for you and directly intervene. You're not gonna get that. And that's something that's really necessary in some contexts. Um, likewise, if you are um accused of a crime and you are in a criminal hearing, you really need a human next to you to help you through it, especially if you are innocent.
SPEAKER_06:Um because of the mental load of the process.
SPEAKER_02:You have the mental load, you don't understand when you need to open your mouth, you don't understand what you can say, what you can't say, how you can say it, you don't have any authority. People think you uh think you already did it, otherwise you wouldn't be there. You're the worst of humans. Yes, and ultimately nobody is ensuring that the criminal system is properly applied except for the lawyer that is defending you in the room. That's the only uh guardian of the proper application of the legal system in the room.
SPEAKER_03:Okay.
SPEAKER_02:Because if the prosecutor just doesn't do the work well enough, or they I mean they're all they're also underwater with cases. So I've seen it already where you try to you try to bundle up a bunch of offenses on a specific person so you can get rid of them, and it doesn't change anything for them anyway, because they did at least two or three of them.
SPEAKER_03:Oh wow.
SPEAKER_02:Um it it does happen. Um and this is not a criticism of the prosecutors because they have a really tough job as well. They're being lied to all day. They have to make sense of really complex or sometimes unavailable evidence. They have a public mission to make sure the society is safe. So I'm not minimizing the importance of the role. It's just that you need somebody to be able to ensure that the that the law is applied fairly and that the person isn't given more time in jail than they should, that they're that the fines are not higher than they should, that they shouldn't be fined if normally you're not fined for that offense. You know, you so that that lawyer is the only person ensuring that the law is applied properly. So that role has to stick with a human.
SPEAKER_03:It has to stick.
SPEAKER_06:Interesting. Okay. I even lost my train of thought. It was very interesting. Yes, I I can I can see your point completely. Yeah, yeah, yeah, exactly. Um, did you see the did you watch the movie uh with Tom Cruise? Uh that he will get people before they committed the crime.
SPEAKER_02:Uh minority report.
SPEAKER_06:Yes, minority report. Good lord.
SPEAKER_02:That is prohibited.
SPEAKER_06:Futuristic.
SPEAKER_02:It's prohibited.
SPEAKER_07:It's prohibited.
SPEAKER_02:The AI Act actively prohibits preemptive policing. You cannot be held guilty for something you haven't done. You can be caught for the offense of preparing to commit a crime, but then it's the crime of preparing to commit a crime. But you cannot be arrested for the crime that you haven't committed.
SPEAKER_03:Yeah.
SPEAKER_02:And you cannot be arrested preemptively because statistically speaking, or probabilistically speaking, you would have committed a crime. That's even worse.
SPEAKER_06:Yeah. And also many times people think about doing something and it never it never happens. They never do it.
SPEAKER_02:Yeah, yeah.
SPEAKER_06:Yeah.
SPEAKER_02:No, the that that is really that's a prohibited AI under the AI Act. With good reason.
SPEAKER_06:With good reason, yes. Okay. The right to digital justice access will be enriched in future constitutions. Oh, sorry. The right to digital justice access will ensure, so we will it will be included in future constitutions.
SPEAKER_02:I think that's true, but I think I don't think they're going to modify the constitution for that unless you need a political stunt.
SPEAKER_03:Okay.
SPEAKER_02:Um, because you can probably already derive that right from existing fundamental guarantees such as the right to a fair trial, such as the right to um to an independent judge, such as the right to information about you. You there there are specific rights that you can use, or the right to due process, that you can use um to construct a right to be able to defend yourself using the most advanced technology. So, in a system where the constitution can be interpreted um in accordance with its time, such as we have here in Switzerland, it could possibly already be done. And I'm not a constitutional scholar by any stretch of the mind, but I think you could do it. Okay. At least in the future, you will be able to. In the US, it wouldn't be possible because they interpret the constitution as their framers intended it. So yeah.
SPEAKER_06:It's different according to a jurisdiction. But the the medium changes, but the core of the principle stays the same. Access to justice is still the same. It doesn't matter if it's access digital or access physical.
SPEAKER_02:Exactly. Yeah. But in a case, but if we're talking about specifically um the right to access an online court system with digital enhancements, that is a more advanced um, that is a more advanced concept. And you you yeah, you probably might need a bit of legislation to get there.
SPEAKER_03:Yeah.
SPEAKER_02:But also it's not necessary to enshrine it in the constitution because you can you could basically modify the existing laws because that would derive from existing protections and guarantees, I think.
SPEAKER_06:You take the principle and then you make it practical with the law. Yeah, makes sense. Okay. You receive an automated notification before you even realize you've broken the law.
SPEAKER_02:Again, illegal.
SPEAKER_07:No, don't even you don't even use the panelists like no. It doesn't deserve it.
SPEAKER_02:And it's I mean, the implication of you being monitored to the level that there is so many layers of technologies and so many monitoring tools that need to be in place for that to happen. I mean, I wouldn't want to live in a society where that is possible.
SPEAKER_06:No freedom.
SPEAKER_02:I mean, that means you're constantly monitored, even when you go to the bathroom.
SPEAKER_06:Oh god. Oh, big big brother.
SPEAKER_02:And the worst level of big brother.
SPEAKER_06:Okay.
SPEAKER_02:It would be a worst kind of big brother.
SPEAKER_07:It's it's not just cameras.
SPEAKER_02:This is you have the drone army that's going to just fly through the window and you need you to do what they want, you know.
SPEAKER_06:Okay, next question: legal time travel. Tech will simulate how a past ruling will play.
SPEAKER_02:That's futuristic and inadmissible.
SPEAKER_06:Okay. Again, in the same realm.
SPEAKER_02:Because you would be it's a reconstitution, but the whole court system is based on the fact that you cannot know exactly what happens. And any kind of reconstruction, which we already do today, by the way, is always subject to variables that you haven't been able to account for. So legal reconstruction, even through AI, would not be um evidence. Like it wouldn't be overwhelming evidence. It would just be okay, this is this is what we think happens, but you would still have the uh the ability as a defense lawyer to raise a reasonable doubt as to whether this reconstruction is accurate or not. So time travel, well, we know it's not possible in the past, at least by today's technology. Um, and it probably would create so much mess that it wouldn't be realistic.
SPEAKER_07:We will have too many multiple universes then.
SPEAKER_02:Oh my god, I would actually love us to invent time travel to the past, but it would be such a mess.
SPEAKER_06:It would be such a mess. People would not respect the boundaries, they will mess up everything, they would like to change everything.
SPEAKER_02:That's really a line you shouldn't cross by that means.
SPEAKER_06:It's interesting to be an observant of the past, but traveling to the past is a different thing.
SPEAKER_02:Yeah, or if you travel to the past, go visit the dinosaurs instead.
SPEAKER_07:Very far away. But then, but then you you crush a leaf and then humanity changes forever.
SPEAKER_02:The great extinctions. So the the the the leaf wouldn't prevent the the the asteroid from colliding with the earth.
SPEAKER_06:But you know that that's a theory, right?
SPEAKER_02:Yeah, the the butterfly theory?
SPEAKER_06:No, no, the the ask the asteroid destroying the dinosaur is a theory. It's not an absolute fact. Well, I mean I I learned about that recently. I I thought it was an absolute fact that there's a contributing factor, yes, to be specific.
SPEAKER_02:You're right.
SPEAKER_06:And there's this the then the the um the dust that happened, the chain of uh volcanoes, all of that is the they have strong uh facts that strong uh indices that it happened, but they're not a hundred percent sure, of course. Yes, so it's no it's not a loan factor, you're absolutely right. But I I yeah I went completely sideways, but it's it's something to be fair, yeah.
SPEAKER_02:Yes, a hundred percent.
SPEAKER_06:It's not it's not the full explanation again with the reconstruction, it you cannot really reconstruct exactly what happened because you you can have an idea, you can have like the some um information, but you don't have the whole story.
SPEAKER_02:Yeah, because some things just don't leave traces.
SPEAKER_06:Yeah, it's true. Interesting. Okay, then justice system will allow crowdfunded berries overrides uh through civic voting platforms.
SPEAKER_01:Hell no. No, no, no, no, this is nonsense. No, not even, no, no, not even. You don't deserve nothing. To be fair, it already happens, but not in state court court. No legally.
SPEAKER_02:It doesn't happen legally, so we have internet tribunals and stuff like that uh to judge stuff on a you know on a fun basis or on a slightly weird basis, but the court of public opinion. No, you don't do that, no, because there is no guarantee of fundamental rights in a court in uh in the court of public opinion, so it effectively destroys what we've been so so um engaged in building for so many years to make sure we don't condemn innocent people and to make sure that we we act reasonably towards offenders.
SPEAKER_06:Yeah, and just because someone is unpopular doesn't mean that they're wrong.
SPEAKER_02:Oh, definitely not.
SPEAKER_06:You can be unpopular and you can be right or innocent.
SPEAKER_02:And and the problem with that is the more you resort to the public opinion court, the more you can damage somebody's reputation accidentally, and that can cause really big issues. You have some people who just really come after you because they think and they may have really good reasons to think it's just that it's not true. They think that you've committed the offense and they just go after you.
SPEAKER_06:They have convinced themselves.
SPEAKER_02:Yeah.
SPEAKER_06:Yeah.
SPEAKER_02:And you can't even blame them because the reasoning isn't that I mean, the reasoning is wrong, but the elements, those are the elements that were given.
SPEAKER_03:Yeah, of course.
SPEAKER_02:So it's a conclusion that unfortunately is wrong, but it's a really hard case to handle then, because you have a person who's persuaded to be correct, yeah, who is relying on tangible evidence, uh, who doesn't interpret them the right way, maybe, who's missing elements, maybe, but it's a valid foundation. And you're just throwing that person to the sharks without a valid reason.
SPEAKER_06:Yeah, just because uh you understand it's right. That's why we need the court. That's why we need neutrality.
SPEAKER_02:Yeah, and we and we don't want the court of public opinion.
SPEAKER_06:No, okay. Final question: Judges will be banned from making rulings without AI.
SPEAKER_02:No.
SPEAKER_07:Don't there, there isn't no you want a no, he wants a no palette. No side. There's no reason to do that.
SPEAKER_02:Okay, because you want the control to remain with the humans. That means you want a human to be able to overrule the AI, or you want the human to be able to make a decision without the AI, potentially on a case that never presented itself to the AI.
SPEAKER_05:Yeah.
SPEAKER_02:And where it's not able to give a good so you can have good practices, which is to do research with AI. That is already something. That is already a thing where you could arguably claim that if you're a lawyer, you should be using all the tools available to ensure the best defense possible of your clients. And the the American uh Bar Association's uh guidelines require the lawyers to already have an understanding of the technologies so that they can leverage them where necessary to help the clients um to the best of their abilities, which is our obligation. We have to do the best we can. Of course. So you can claim that it's already implicitly necessary. I don't think it's been ruled on, at least in Switzerland, but it could already be there potentially. Um but you wouldn't find you you wouldn't have a rule that says the lawyer cannot do anything without an AI. Because then potentially, if you have a disruption of electricity, you don't have an AI anymore. You don't have a court system. And there's nothing to do then. You would just sit and wait.
SPEAKER_07:Yeah, we just have a holiday.
SPEAKER_02:And people can wait for their case. That would be terrible, terrible.
SPEAKER_07:Yeah.
SPEAKER_02:That would be beach level judgments.
SPEAKER_06:Okay, so let's go uh to the final big picture question. If you can give one piece of advice to people building legal systems, platforms, or policies around AI and justice, what should they never forget? Especially if the goal is fairness for everyone.
SPEAKER_02:I think I have to go back to my previous points to policymakers. I think. I think it really has to be making sure that you're not depending on any single actor, that you have redundancies, that you have security, that you have ethics, that you have economic intelligence, even because you're going to be a target of hacking attempts, you're going to be a target of so many different things. Um, you're going to be, you're you're you're going to want to be able to respond to different kinds of scenarios. So make sure that whatever it is you're building, when we're talking about online courts and using AI to help people settle cases, you have to make sure that what you're building is as resilient as it can be. Um, and if we're only talking about AI creators, I don't know. I mean, everything's changing so fast. I it's just like you're you're the moment you come out with a software, somebody comes out with a more modern version of your software. It's extremely frustrating. My really big complaint about AI is that it's going so fast. We do not need to pour the trillions of dollars that we're currently pouring on AI. We don't need to do it. I really think that we should take a step back, like Elon Musk advocated, until he had caught up in the race with everybody else and then he stopped advocating for it. But pause and try to develop these frontier models, so the most advanced models, try to develop them in a context such as CERN for uh nuclear particle research, for particle research, such as ITER in France for fusion, uh fusion energy, build a common science lab. Um, pour those trillion dollars down there and cooperate because that reduces the incentive to engage in a war, in an arms race.
SPEAKER_03:Okay.
SPEAKER_02:And second, have institutions like the the atomic energy um, the institution, the the agency for um, I forgot what the acronym is, but the international agency for atomic energy, um, I got it. Um, use something like that to monitor, to check what everybody's doing so that you have guarantees that while you're holding back from developing these really dangerous technologies, the other ones are not steam, steaming ahead and overtaking you.
SPEAKER_03:Okay.
SPEAKER_02:I think that's really that's really what I would be focused on is trying to cool down the race, build institutions so that we build it reliably, like we're doing for other frontier technology, and try to have an organism that can help us work together. Try to have treaties like the denuclearization treaty. Um it was successful until it wasn't, but at least it was successful in reducing the stockpile of nuclear weapons. Um I'm very happy that we have six or 12,000 less nuclear warheads in the world. Um I think you have to follow the same kind of logic because the amount of money we're investing now in AI is incomparably greater than what we invested to develop nuclear weapons.
SPEAKER_07:Okay.
SPEAKER_02:And we still have a climate change challenge, we still have educational challenge, we still have fo uh hunger in the world challenges, we have a, I mean, we have so many challenges. Uh, we don't know how to how we're gonna deal with the next pandemic. If you know we haven't changed our habits. So if human experience is any kind of guideline, it will happen again because we haven't changed the behaviors that led us to that.
SPEAKER_03:Yeah.
SPEAKER_02:So yeah, I think that's the that's my closing word. Let's do this responsibly, using other models that we've developed to do it responsibly, and that actually worked.
SPEAKER_06:And keep the human always.
SPEAKER_02:Oh, yeah, you have to keep the human in mind, of course. Um thank you. This has to be done by humans anyway. Um there is no there is no future for humankind at the moment if we continue on the path we're on, because the odds are that we're gonna mess up really badly.
SPEAKER_06:Oh good lord.
SPEAKER_02:So that's the gloom and doom to end the talk, is it?
SPEAKER_06:Yeah, beautiful. So justice doesn't have to stay locked behind long waits, high fees, and legal gagund. It can be translated, automated, and still holds its sold. Thank you, Nicholas, for showing us how technology can open more just than apps and it can open systems that have felt closed by too many for too long.
SPEAKER_02:Thanks to you. It was a real pleasure to chat with you and thanks so much for your insights. It was a really pleasant chat.
SPEAKER_06:I hope that we'll get too less intercambios. There's always a little light.
unknown:Thank you.
SPEAKER_00:Thank you for listening to Intangible, the podcast of Intangible Love. Let's talk about intellectual property. Did you like what we talked today? Please share with your network. Do you want to learn more about intellectual property? Subscribe to one of your favorite podcasts players. Follow us on Instagram, Facebook, LinkedIn, and Twitter.
