Let's go. Let's do it. Thank you so much. So as Anna put it, I mean, she basically framed the discussion around why AI and ethics matters.That's really, the the whole, discussion here. But I also call it making the case for humanity. And if we really want it to be provocative, is it still relevant? So, one of the things that we have to think about, as we talk about this particular narrative is there's a positive and we could say the peril. There's the, you know, there's something that's could be beneficial here when we're talking about AI. There could be sort of this dystopian view when we're talking about AI. It's, we're now we're we see it as rapidly, really much, transforming our world, but there's there's opportunities, but we have to think about what are the ethical issues surrounding, surrounding it. You know, what are what are the issues that when we're talking about the future of of humanity?

And the one question that we need to, ask ourselves is where does, accountability lie? I think this is so, so, so critical and important. You know, when we define ethics, we kinda define them in sets of principles that are very, very common. They're they around transparency, fairness, accountability, privacy or privacy, and what's beneficial and what's probably non, beneficial here. And so when we talk about transparency, it's gotta be open. We we have to think about that is how do these systems work? And there are lots of courses on describing that. Where's the where's the fairness in terms of the discussion itself, you know, in terms of where potential for bias can exist. And we know they do. And of course, how does, where does accountability lie, and when we're talking about particular outcomes of these AI systems.

And I look at it as as a systems approach here, in our discussion. Privacy perhaps is probably one of the most, important issues and topics. You know, who what and who and for whom are data being gathered. And so it's looking at, you know, protecting what we call the individual rights, and that's gonna have dependency upon laws and jurisdictions in particular, countries or and, and also states and so on depending on where you live. And as we said, you know, from the beginning and topic, from the beginning of this whole discussion is what are the benefits and how can they benefit humanity? Now we have Jeffrey Hinton here who who's sort of the, we could say, one of the godfathers, if you will, founders of of what we talk talk about AI, formerly of Google, Sam Altman of of, you know, the whole Sam Altman particularly around, you know, the whole c e he was the CEO when we talk about, OpenAI.

But I also wanna say that there are women, who are very, very key here. I haven't I had not put their pictures in only because, of space. But the women, since this is a women's con tech conference, include perhaps two very, very important key women in this area, doctor Feifei Li, who is at Stanford University, very, very premier in what she's been doing and looking at the assessment of AI and what is AI. You know, how do we deal with AI when it's, for good? And particularly, also, I could also state, Mira Mira Mirati, who was actually the, former CTO of OpenAI. And she's now a CEO, at, Thinking Machines Labs. So this is kind of the area in terms of how we define the ethics, the people behind it, key people, key leaders behind it, but something that we have to think, keep in mind when we're talking about these particular issues. However, one of the top concerns are around the the biases, and there are various biases that we can imagine.

It can basically, when we're talking about a and we're talking about learning, language models, if we double click further, you know, it's we we think about how the amplification of societal biases are are made and and and developed. And and, basically, it's these are societal discussions if you can, can imagine when we're talking about learning language models. Yeah. Some of the issues that we have to be very, very cognizant of is the use of facial, recognition technology. There's a corpus of research around where there are errors in terms of how these, facial recognition technologies are used and the biases that can be can be found, in the usage of the, facial recognition technology sets. Hiring algorithms all also also can also, play an area in terms of the biases behind hiring. Often used by companies today, But are they missing really the human, contact or the human, intervention, if you will, or the human dialogue?

I I think this is something that we have to or questions that we perhaps have to ask ourselves, especially then we can get into, approvals for loans depending on, again, the buy there are biases that are we have to be cognizant about, on, especially when we're looking at how algorithms are used to approve loans, and then the criminal justice system overall.

So what the implication here is we do need to have interdisciplin interdisciplinary teams, to look at how we mitigate biases. And that goes for the development and implementation, if you will, of, the use of AI and AI systems, learning language models. And the area and one question to keep in mind when I, framed up the discussion is who is accountable? You know, what does safety look like? These are kinds of the concerns that we have to be very, very cognizant of when we're thinking about these sets of questions and and issues overall. There is something that we have to really, really, really pay attention to, and that is the importance of transparency. Remember I talked about or I mentioned one of the five principles over here is transparency.

And if we have what we call AI systems that are simply looking like a black box, you don't see what's going well, what's coming on, going in, It makes it diff difficult, really difficult and challenging to actually identify the sets of biases that we and errors. And, actually, how do we correct these errors? So we do need to think about what is often termed as explainable AI. And so explainable AI has to be it has to be clear. It has to be explainable to to humans. It cannot be something that is very scientific in terms of as lingua franca, and people have to understand what it is your one is talking about when when one is using these types of, systems. And then, of course, what we have to think about when we're talking about explainable AI is to think about, you know, what we could get, or obtain as a result of its, of explainability, and that is trust and, and accountability. And we could also think about, you know, optimizing decision making, when we're thinking about the usage overall.

So the importance of transparency in this model discussion or this modality of discussion is very, very key when we're thinking about AI, ethics, and, of course, looking at the usage for humanity. The area that becomes of great, great concern is around privacy or privacy in the age of AI. You know, I asked and posed the question of again, one of the concerns is or one of the principles that we, identified is how are data gathered by whom, by which entity, and for what purpose. And so there's a variety amounts of data that's gathered. Again, double clicking on, large learning language models could, really, really raise concerns about privacy overall. And kind of a a notation here is potential for surveillance. And so there is this tension between innovation or I'll call it polarity between innovation and individual rights between, also polarity by companies and organizations to want to gain rev gain quick revenue.

And again, where is the about, boundary for protection of individual rights or and or do the user or do the the individuals know what's what the data how the data is used and for what purpose? Now there are regulations in some, countries or jurisdictions. Of course, we have if you were looking at Europe as the data protection regulations are known by its, acronym as GDPR. There are there's the use of privacy enhancing, types of, technologies to actually, kind of protect be use the, for protection of your privacy. And, of course, understanding the ethical implications, if you will, of of gathering the this data. And I think this is a kind of a a a this is a very major concern because what happens is that, you'll we could hear that companies organizations are not gathering data.

They're simply, you'll have lines and reams and reams and reams of what we call what I'll call legal explanations. Especially if one wants to be able to use a particular product or whatever. But at the end of the day, privacy is is can be of a concern depending on, where you're coming from. And the other issue around privacy could be and this types of although tangentially related is also the use of, deepfakes. And we know that that's a very major issue, and there's, you know, there's laws that are currently being passed about the the use of these types of deep fakes. And, you know, again, your privacy how how we can imagine how privacy can be really, really transgressed if we're not careful.

So, this is another area that, if one were to ask me the question, what keeps me up late at night, that's probably one major area. The future of work, you know, again, it's not it doesn't have to be dystopian, but there is, there is a concern. People there are organization concerns that people said, whatever it's automated will be automated. This is just the reality. And so, there is the narrative, oh my goodness, this could displace our workers, this could displace people in terms of jobs, This could lead to, economic unrest, or disruption and social unrest. But there are opportunities for new jobs here. So on one hand, you know, again, we're at this sort of we really are at the fork in this road of the what could be the possible of, really the the possible use of these technologies to create jobs, to create new potential for innovation, but also, being able to pivot into those new positions.

Because this this has been a reality for the past several years. We're just beginning to become you know, take note of it. It could be use of, you know, looking at how we, invest in education training for people who are in danger of this automation or, most, you know, danger of what automation be in terms of job displacement. They're in creating, you know, what we call this safety net for social, for for society. And then, of course, what's very, very important in my narrative, it has has been for overall is really looking at the human centered approach to any of the technologies that we we, we look at in terms of automation. The human in the loop, the human in the loop. Some there is an argument to be made or polarity that says the human doesn't have to be in the loop.

But again, in my, I would actually argue the contra. The human actually needs to be in the loop because human is actually creating these sets of technologies for the use of, however, you know, the automation for work. So, you know, you could look at, getting certified in air certain areas, you know, take ownership of your own career in terms of what this could mean for you. I do walk my own talk and, I like to understand what I don't know, I don't know. And, I am, looking at, several types of certification in the particular in the in this space. When we think about human values, again, we talk about flourishing. It's gotta be around humanity. Remember, that's the title about how would he how do we talk about humanity.

And that's really important of looking at what does public disclosure look like, what does an ethical framework look like, and, you know, what we have these interdisciplinary groups of individuals and leaders coming together with a background to create these models for ethics, To, look at what it means to work with policymakers to, you know, it's not about regulation for regulation's sake.

We need to take regulators along a journey together. This is extremely important because, otherwise, we can over rotate on the other end, and we have we stifle the potential for the what I will call the benefits, of of AI overall. There are, a plethora of case studies that we could imagine, but they do, you know, they do create the ethical dilemmas. So they're very commonly used, whether we're talking about autonomous autonomous vehicles, the accident scenarios, lots and lots of cases around that. Health care decision making, you know, what are what kinds of decisions are being made when we're using these technologies for health care? Are we cognizant as patients or as customers of the use? And should we be? Well, the argument of of course, is yes.

And then, of course, the, the impact, perhaps on these sets of technologies around democracy and and the freedom of expression, going back to how privacy can be, aggressed, if you will, or transgressed. So I will say here, I do have a call to action, as this is a a keynote. We all have a role to play in terms of shaping ethical, you know, development of AI. And it's asking the question. It's constantly posing questions around safety. What does safety look like? What does accountability look like? So for individuals, it's really train and educate yourself. Be cognizant of what it is. Engage in the conversation. I think this is extremely important. Organizations, we need to adopt ethical AI principles, as organizations or companies. And where they lie, it depends, you know, if a CEO is not caring about it, then it's not important.

If it's not on the CEO's radar or or looking at, looking at it from a reporting perspective and also kind of balancing it from a business perspective, then it's perceived as not important. Otherwise, people are pay playing lip lip service to it. Again, thinking about transparency and fairness and, of course, posing the question as to where where does accountability lie. And for policy makers, it's frameworks, you know, looking at how we develop these ethical frameworks, and this gets into safety for citizens. Who's responsible, at the end of the day? Who shares this accountability? And I say to all of you, let's work together to ensure that AI benefits are, you know, are for humanity. And that's really very gonna be very key for our society.

So with that, I leave you with these thoughts, and I really hope that you, concern yourselves with asking the cook, these particular questions. It is around about the art of the possible. It doesn't have to be a dystopian view, but on the other hand, we have to think about where the shared accountability lies for the sake of our humanity. Thank you.