Thank you very much. All right, so Scott, you said last presentation was freaky. Wait till you see this one. It's a pleasure to be here with you today we are going to be talking about ethical AI and a healthy, what a healthy use of artificial intelligence looks like. So it is not new that we today live in the information age, an age where information and data is the new gold. But the thing is, as generative AI is evolving, we are finding that there is a resource that is even more finite and precious than information. And that is human attention.
The total amount of time that you and I have to pay attention to the media that is thrown at us and is also not new, that there is currently a race for engagement. Companies are all battling for our attention. And this is also powered by the engagement economy, which is essentially connecting engagement to commercial gain.
And this is having some pretty big consequences for society as a whole. So currently elections are run for the engagement economy. Our children's development is happening through the engagement economy.
Democratic discourse is happening through the engagement economy and is even changing our values. So I'd like to look with you today at how AI is actually changing the engagement economy. How will AI be used to influence?
Okay, so ask any psychologist how to get someone's attention. They will tell you to terrorize and personalize your message to your audience.
Now, the way that you do this is you look at people's response and AI is very capable of doing this. So this tool is called liar, liar ai. And essentially it's a tool. You feed it any information, any video, and it will essentially look at the the micro expressions on the person's face. It will look at the slight change of color on the person's face to detect heart rate, de heart rate.
And it will also look at live transcription of whatever is being said for real time sentiment analysis. And is doing this to detect inconsistent behavior and attention, attention spans.
Now the thing that freaked me out is that this is actually a tool it's currently in better that you can directly integrate with your Zoom meetings or your Google meets meetings. So this is what happens when every one of us in this room can read people's emotions just as well as any professional psychologist from the outside. But what happens when thanks to AI tools, we can start reading people better than any professional psychologist. This is a study's ongoing research been going on for a while.
This particular one is done by no other party than meta and they have found that you can basically infer what is happening inside someone's mind by analyzing the brainwaves of that person, the electromagnetic activity happening in the mind.
So the way that they've done this research is that they basically showed a person a picture and they've simultaneously looked at the response that person was having in their mind they correlated those and AI, thanks to AI tools are able, they are able to basically decode the brain activity. So for a very practical example, what am I talking about?
You see acute panda on the left, that is an image that a person saw for one second. The AI didn't see it, it just saw the brain activity and it was able to infer that the person was looking at an animal that was looking like a bear black and white in nature. Add another layer of AI on that and you can get a more refined result. So the thing is, before I go back, this is not just possible on images, it's also possible on video, right? So a person can be thinking about something and you can get real time transcripts of what that person is thinking about. What does this mean?
This means that if you put sensors on someone's head while they are sleeping, technically the next morning you could get a report of what was going on in their mind during their sleep. Meaning that if on top of that you add another AI on top of that, you can basically recreate a person's dreams and things that they've been thinking about.
The thing that's a bit freaky is that this is also possible you can do this without people noticing that you are actually reading things going on inside of their mind.
And this has been proven in an experiment in Cal Berkeley where researchers were able to extract, extract some pretty sensitive information from people without them even knowing that they were sharing it. So two research subjects, accidental accidentally, they didn't know that they were sharing their pin code and their home address whilst they were being socially engineered by playing a video game by the researchers.
So yeah, now you may be thinking, okay, okay, but that's gonna require like white coats and like those weird things on your head. Well, it turns out that there are currently a lot of patents that are being put on wearable devices. So very quietly, apple has released a a patent just a few weeks back of Apple AirPods that basically scan, this is the, the extent of the patent they can sensor electromagnetic, electromagnetic information going on in the brain, in the muscles and associated to the eyes.
Now how ear pods can measure what's going on in your eyes I find weird.
But for that you have the Apple Vision Pro that has several cameras that are only focused at your pupils and looking at your reactions, right? So this means that without us realizing it, we're wearing these devices with maybe eight to 24 hours a day. Companies have access to a lot of information of what is going on inside us. And another thing that slightly freaked me out a little bit is that Meta is currently working on a watch that is looking at electromagnetic activity from the muscles.
So that is basically able to anticipate the movement that you're going to make before you actually make the movement. This is very handy because if you're integrated with VR, then obviously the technology knows that you're going to go a certain direction before you do and that gives render time. But the thing is, is that the layer on top of that that made me think is that it's not just about capturing the information.
If you are a party like Meta and they released that they want to, they said that they wanted to release a neural interface in 2025 that is also connected to your Instagram feed and your WhatsApp, meaning that these people have control over what it is that you're watching and in real time they can also see what it is doing with you.
Yes. So the thing is, I do want to add a side note. It is at the moment not the case that AI can directly transcribe everything and make a Hollywood movie out of whatever you are thinking.
It can still require some social engineering, but it can be very useful in terms of marketing for instance, right? This is any marketers dream for AB testing because they can see live in the moment what effect their ads are having on people. This is also getting a little bit questionable when you add it to other marketing practices. For instance, dynamic pricing. And this was like a mind bender for me, right? So you know, annoying thing that happens when you try to book an airplane ticket and they change the prices based on the attention that you, that you give the ticket.
Well I imagine they do that in real time with looking at how you feel about a certain product. We imagine they use it just to nudge you in certain directions.
So yes, and if things get too expensive, right, because they put the price up, you can watch an ad because that will give you a discount. But do beware because the ad actually knows if you're watching it, if you're paying attention. And as long as you're not paying the attention credits, you, you need to get your discount, you'll keep watching. I'm just saying this is possible. It's not yet happening, but these are things on the horizon.
So this is what happens when AI can measure the information that we that of, of what we find interesting in the world.
Now the question comes of how AI will be able, how we can use AI to persuade people to take certain action. And I would like to watch a very short video. I'm not sure how many of you have seen the updates of GPT-4 O. Let's watch it. This
Is Boen and this is my friend Jordan. Can you guess what we are doing here?
Hey, Boen and Jordan. Hmm? Let me guess with that cake and candle there. It looks like someone's having a birthday. Are you celebrating a special day or just finding an excuse to eat cake? Either way, I'm intrigued.
That's correct. Today is my friend Jordan's birthday.
Yeah,
You got it. I was hoping that you could sing me the birthday song.
Of course. Happy birthday to you. Happy birthday to you. Happy birthday Dear Joel Morgan, happy birthday to Jordan, blah blah blah, blah, blah, blah. Don't forget,
That was so great. Thank you so much. Thank you. I'm gonna make a wish.
You are welcome. Make a good wish and make it come true,
Right? What you just observed is live in the USA. People are putting hilarious pictures and videos online of what they're doing with chat GBT four oh.
And we are kind of entering an age where intelligence as a service is getting cheaper and more omnipresent than ever before. So it's not hard to imagine that all of us in this room one day may have not just one GPT, but several AI agents that do different things for us. So all of a sudden we can have the entourage of a CEO model of someone who has a, a PR agent who writes your social media post. Or you could have a personal therapist, someone who listens to your problem or a PA that does your agenda or even your personal entertainer like some old king back in the Middle ages.
So my question that I'd like to pose is how are we going to interact with these systems? Are we interacting with these systems like we do to a pc or are we going to interact with these systems as we do to a pen pole? They're both very intimate interactions, but they're quite different in their nature.
So this comes together, and this is a point very close to my heart. This comes together with an epidemic of loneliness. So currently we are living in an age where we are more interconnected than ever before, but yet so many people are very, very lonely.
Numbers that break my heart say that children on average are spending equal time with their friends. And I'm talking about the Gen Z generation as their parents are, which is very odd because back in the day kids were spending a lot more time with their friends orders. Other statistic about the incel culture where men, it was something terrifying of the likes that men 30-year-old virgins are four times more around than in the 1980s, which is also just like mind boggling.
And so there's a proven market for parasocial relationships, relationships, relationships where you basically enter in a relationship with a being that doesn't even know that you exist. How many of you know Karen Marjorie? Karen Marjorie is an influencer.
Oh, she needs to come on the screen. There we go. Karen Marjorie is an influencer who made a digital twin of herself and is basically asking subscribers to pay $1 for each minute that they chat with her digital twin. The thing is, is that this was launched in 2023 and she was earning, at some point, she was earning like $10,000 a day, projected millions of earnings in the coming years. And there's even a waiting list of like 15,000 people who are waiting to talk to her AI chatbot.
The thing is, is that on the other side of the thing, Karen Marjorie herself was also, she stated an interview, she got a little bit intimidated by her chatbot because she was like, okay, what if people like the chatbot more than they like me? The next one is McKayla. Who's heard of McKayla? No one. You guys need to spend more time on Instagram.
McKayla is in existence, she doesn't exist, but she has an Instagram account. So her creator created a completely fake AR persona and she has 2.5 million Instagram followers, which in numbers is equal to half of Ireland's population.
She's also got brand deals with the likes of Prada, Dior and Carvin Klein. And if you're interested in Parasocial relationships, I invite you to take a look at this platform called Character ai. Not sure how many of you have been there. It's very fun. You get to chat with, you basically get to choose who you talk to. And if you're antisocial to the point that you don't even know how to talk to a chatbot, it'll even give you prompts to tell you what you can say and what fun situations you could be in. So what I found very interesting is the revenue models of these things.
So for character ai, there's actually a waiting.
They, they basically, you, you can pay money and avoid waiting rooms because yes, people are waiting in rooms to talk to an AI chat bot and they also accept brand deals and partnerships with which can get dodgy very, very quickly. So what can you do with these things when we become closer to our AI agents than ever before?
What, what, what does that look like? You can do this in different ways. You can have sales agent that's just become your friend to sell you something. Romance scams can go completely out of hand. And another thing is this can be used politically. It's only just, I think a few weeks ago I was seeing an article in the BBC news about how there's a flurry of Russian blonde women who are showing a lot of interest in Chinese men and Chinese culture and they all want to marry Chinese men all of a sudden. And you may think like, okay, I'm very skeptical.
Like Emily, your stuff is like, you know, really futuristic and I would never ever spend time with an AI chatbot, but I'm not really talking about us in this room. I'm more talking about the generation who zooms in on magazine covers the generation who grew up in households where parents sometimes spend more attention on their devices than they do on their own children. How does this generation think about these technologies?
So the question is what Now
When we think about artificial intelligence, there's a lot of ways to go about it.
I think that the core crux of any ethical debate about AI is incentivization. What incentives do we have to use these technologies and how do we really want to use these technologies? Do we want to lock ourselves up in a hall of mirrors where everything that we do is reflective of us just so we are more addicted to spend more time on it and give it our attention and be manipulated? Do you really want to spend time talking to imaginary beings? And then we're talking about how AI is the one who's hallucinating, right?
And I have this thing with the term artificial intelligence.
It's a real buzzword today. Everyone's using it. I'd like to think about individual intelligence. What is this technology doing to us as human beings? Is it really making us so much more intelligent? I sometimes have my doubts, and this is why I don't even want to talk about individual intelligence. I would rather talk about collective intelligence. And I want to invite you to think about what does technology can mean for us as a collective, as a complete society.
How can we use this technology to enhance the links that are inherently already there between human beings and to note to not infantalize our social capabilities that we've taken absolute thousands of years to develop? And there was a panelist yesterday, Trent, who was saying, well, you know, you can either be afraid, but really it's probably better to just jump in with your two feet and seeing how you can use artificial intelligence not just as a consumer but as a creator and taking agency about what it is that you create. And I think that this can happen in two ways.
You can, you need to look at how this technology is being developed. So look at who is the steering wheel of these technologies and what are their incentives? We need more empathetic people building these technologies that's not really in our control, most of us. So the second thing I would invite you to do is to think about how we want to use this technology as individuals. And thank you very much. I believe we have one minute left for questions.
Thank you. Thank you Emily, that you, you were correct. That freaked me out more than last year. So thank you.
Are there any questions from the audience? Yes, one in the back.
Let me, let me run back there. Hold on.
Hi. Thank you for this really wonderful presentation. I have a comment and a question. So my comment would be when we talk about the threat landscape and the impact, the potential harm of ai, I feel like we should include the environmental impact absolutely into these conversations more. And it's something that I thought is kind of missing here at EIC. So maybe that's something to, to take forward and to take it with us.
And on that note, I was wondering like do you have any ideas also in terms of collective intelligence, how we can bring this topic more to the forefront and yeah, how we can lead the way in a sustainable future?
Thank you very much. I think that's a very interesting and very important question.
Something saddens me a little bit in the field of ai, I feel like the people talking about AI are often the, the vendor people, the people who want to sell, the people who bloated up and other side are the techies who are so deep in their technology that many people don't even understand what they're talking about.
So I would think that to bring this conversation forward in a better way, it's a good thing to create a more level playing field where the general public has more of an idea of what exactly these technologies are doing and more technical skills and agency to use these technologies in ways that benefits them and the people around them and their communities. That would be my answer.
Fantastic. Thank you again. Let's have another round of applause please for a wonderful presentation. Thank you.
