George: Hello everybody and welcome on the Lights On Data Show. This is episode 1 of season 3. This year, we're going to cover, quantum computing and AI, machine learning, data science, data governance, data management, and so much more. And today's topic is very special to me because of something that I'm very interested in. And I think we're going to just hear a lot more about it and these coming years, and it's “why society needs a AI”.
And to get us started, I'm going to hand over to my co-host, Diana, to introduce our guest speaker, Kye.
Diana: It's my absolute pleasure. Today's guest is chief experience officer and partner at Cannucci. He also established the Swedish AI council and the Nordic AI alliance. He is a business and product strategist or writer, designer, public speaker, brand builder, developer, and AI enthusiast.
He is on a continuous learning and growing dream. He believes in AI and its immense potential to advance humankind, fundamentally impacting health, food, production, energy, business, and creativity. And today we want to hear all about it. Everyone, please welcome Kye.
Kye: Thank you very much Diana. I'm very happy to be here.,
Diana: We're very happy to have you here, Kye. On a personal note, let's start for a little bit of warmup. What do you do for fun in your free time?
Kye: For fun I learn about AI and personal development, but besides that go skiing. I dunno, nature, cooking. I think not anything abnormal. I really read a lot of research and follow the news.
Diana: It's amazing that you can connect what you do for work with fun being that's a great recipe for success. So why AI and not something else?
Kye: I fell into the field of AI, I think by chance. I worked in innovation agencies for most of my life and one of my old creative directors was one of the partners starting up a new fund, a venture capital fund, and he asked me to help on board and help out the new investment they'd done in one of the most interesting, deep learning platform companies in Europe. I hadn't even spent a minute thinking about AI before that moment. And then all of a sudden there was the most exciting field I'd ever encountered. And after two weeks helping out, I said, I want to be in this. This is going to change the world.
George: Do you feel it's changing the world already?
Kye: Yeah, definitely. I would say pretty much every company from startup to large company, that's doing amazing things or having AI at the center of what they're doing. And also, but if we look at politics in countries, both on a macro level between countries, but also inside countries, the exciting things that are happening involve some sort of artificial intelligence or neural network or machine learning. Definitely.
George: When you and I had a chat before to discuss what the topic will be, you're mentioning that you're seeing that democracy is under a lot of strain and it's under a lot of strain because the politicians are just becoming more distant from their constituents and the problems that they're supposed to help fix.
And, media is also turning to be a bit more clickbait than it was in the past. And it's not really scrutinizing politics like it used to be. Of course media should be that third state power to keep politics in check and that's not really happening anymore. Do you feel AI could step into that void?
Kye: Yeah, definitely. I'm hoping. And also we're working on that. I work with the Swedish national AI center now and we're for the last year, we've tried to work closely together with the big media houses, the big newspapers here in Sweden to use AI more, to support journalism, because I think the journalistic profession has been the profession that has been under the most strain for the last 20 years because of internet and digitization. So they're almost pushed into a corner. And when you're pushed into a corner, it's mentally very hard to embrace another new technology on top of that, the one you're currently fighting. In essence AI is what I think journalism could use to take back power, to assess some analyze tons and tons of information in a heartbeat, and go back to quality, journalism and safeguard democracy.
George: Can you give us an example of how AI can help with that? Can help journalists?
Kye: Yeah. Guess what everyone, a lot of the media, the news we see is in one sense or another published or produced by corporations, it's called lobbying or public affairs with a more fancy word.
What AI can do is help us identify what is objective journalism and what has a source and an objective to actually change the truth and affect public opinion. And when we go to emotions and political suggestions, AI can help us cut through that and see what is actually good for the future of society and what is good for corporations and short-term financial gains.
George: That's very interesting. I remember we had a chat with somebody from Thomson Reuters and they were using AI and natural language processing to basically create summaries of all the articles that they were consuming. So they would provide their analysts with just a high-level viewpoint on, Hey, this is what the article is about and you can go more in depth if this is what you're looking for.
Kye: Yeah, I think those models that you are mentioning now, the natural language processing models like Burt and the new ones even more powerful ones that are open source, I think can help us safeguard and reinvent democracy in a sense, if we can have AI models as a supporting layer to journalism. We can make sure that journalism covers not only what's most popular right now, but we can actually assess and see, Hey, are we really looking at the topics that matter at hand? And are we looking at the same things that all of our populations are interested in and not only the cultural influencers.
George: Is this something that you only see happening in Sweden or something globally, that more countries are coming together to share data and to be able to make this happen for their own societies?
Kye: I think we've yet to see a lot of data sharing between countries on that level. I think in the large corporations it's happening in an industry, but not on in the media sphere. As of now I think there's a lot of intellectual property safeguarding still. And from what I see in the U S I would say the big newspapers are increasingly using natural language processing as part of their work process, but that's also because they have a large enough reader base to be able to do that.
I think we're going to see some differences in terms of market size, when it comes to being able to have enough data and infrastructure to work efficiently with AI support generalism.
George: I would also imagine that AI is currently used in the military quite a bit and in the central intelligence agencies to safeguard the safety of the citizens, but as Kate Strachnyi has mentioned here: "what are your top three use cases for AI in society? What other applications are there for AI for the benefit of society?"
Kye: Top three. Okay. I would say one is to change healthcare from today's sick care, waiting until we break down. And then we seek care to preventive maintenance, actually using all the data that we have on us all the time from the accelerometers, in our smart phones, to our smart watches, our pulse data to our heart rate, variability, So today we have all the technology that it would make sense for a doctor or the healthcare system to contact me as soon as my body deviates from my health.
So that's the first thing and that's good for everyone. And then in that, obviously I would include mental health as well. So the state could step up and really help citizens to become healthier and happier. That's the first thing, the second thing I would say is schooling and grammar school, I would say, instead of educating our children on conveyor belts in 150 year old system, we should use AI to at all times better understand what they are prone to learning at that single point in time and provide the teachers with the knowledge. Because kids, they develop at different stages in a different pace and understand different topics and contexts differently.
So I think it's unjust to have them learn the same kind of math at the same time, even though they are differently, different mature. And I would say, I think the third one is true. Where we just had the 26th climate conference, which is just insane in itself. And we're still not seeing real action because we're beating around the bush and avoiding the actual topics. I would see more AI cut through the bullshit and actually address and help us simulate the future based on what's what decisions we make.
George: From the first two, I'm gathering that it will be a little bit more personalized experience for our society at a healthcare level, at the educational level, it's something that's catered to me as the individual and not me part of the masses. And of course the second one, the truth, that's all prevalent. It's something that we need constantly. And of course, within the climate change in particular. Have you seen any headways been made in any of these?
Kye: Well, in the first two, definitely that's happening. There's a lot of companies and a lot of initiatives. We are working with the healthcare system in Sweden and already the big health care providers in the US so we're increasingly using AI and in China as well, I know. So that's happening. In education, I think the biggest business gains are still in adult learning. But I think countries like Israel and like more progressive countries that understand how everything connects, they will jump on board very soon. And the last case, not yet, unfortunately. it's still so uncomfortable to change and so very comfortable to just keep discussing matters since they've actually changing.
George: We have a question here from Petrie, who's your neighbor in Finland. And he's wondering how is the government funding situation in Sweden for AI, and also, are you involved with EU?
Kye: It's fun that comes from Finland. I would say that we're okay funded in Sweden, but not as well as Finland. Finland is probably a hundred times more funded than we are, and that's also why they're making a lot more progress and moving faster ahead. We have slightly different models with the more decentralized model here in Sweden. But it's okay, but it could definitely be more. It is a numbers game, but both in terms of money and access to data, obviously if you're MIT and you can put 1 billion into AI research, you're going to make more progress or like in Canada. And yeah, we are talking to the EU parliamentarians and the EU and actively involved in trying to push what's happening.
And I think we've made a lot of progress over the last four years. But it's a little bit of a struggle. EU has closer to regulations than actually putting concrete goals on the map.
Diana: I have a question regarding the future of AI. So right now, do you see any changes in the education that prepares people to will work in AI in the future?
Do you see that universities are moving towards preparing better professionals to advance AI? And related to this question, do you think we should do more when it comes to educating people, educating professionals?
Kye: That's an interesting question. I think most universities around the world have stepped up a lot over the last two years in terms of their curriculum and how they educate, but an even bigger change has happened in companies.
And even in the legacy companies where we see less emphasis on actual degrees, it's much more natural. You can hire an high school student, as long as they understand the tech and the tools. You don't have to have an engineering degree. Obviously, if you want a researcher, then it's probably good to hire that.
But if you want someone building and doing, it's not necessarily the same thing. You can learn. And I think the tools are progressing so, so fast. So it's not necessarily that if you spend four and a half years getting your degree that's going to be valid anymore.
George: As Ravit is mentioning "very interesting topic", and I do encourage everybody to follow Kye Andersson on LinkedIn and actually check out some of his videos on YouTube. And he's a great speaker, a lot of conferences. So if you ever get a chance to listen to him, from what I've seen on the YouTube videos, that's a talk that you want to go and attend.
So we have another question here from a Petri, Kye: "I think we are still under investing, and it's not coherent. Happy to talk. Thanks. Clear answer."
Kye: Yeah, no I would like to bridge from Petri's comment there. Yes. Where the levels of investments are really low compared to the value potential.
And that is a signal that those we have elected to safeguard our future, or not maybe fully understanding what value this holds or whatever.
George: And it does feel like China, for example, is investing heavily or so much more than a lot of the Western countries.
Kye: Yeah, definitely. But they also have the freedom to not be, have to change government every four years.
George: Right. It's easy for them just to make a decision.
Kye: Yeah. And it's, if I would encourage everyone to read China's 2030 plan, it's a very interesting read in terms of how they've set concrete goals, according to the UN industry standard and what every industry should be able to do thanks to AI in 2030, and then they've designed their entire economic incentive structure around that.
Diana: Any other resources that you suggest, Kye, to our listeners?
Kye: Wow. Resources. I would say a lot of therapy and organizational development are probably number one to turn, even before you go into learning and understanding AI.
George: We would also like to encourage them to check out AI.se.
And can you tell us a little bit more about this, about AI.se?
Kye: I work part time at the Swedish national AI center, AI Sweden, which is it's government funded, but it's also funded through the Nordic model system. So it's part funded by the government and part by corporate partners. And we have a couple of hundred companies as partners now across all of Sweden because everyone realizes most of the challenges we have are too big for anyone to solve by themselves, apart from AB Erickson and the likes. So we share data, we work on different models. We train like Swedish medical language models, part together. I think that's like a collaboration between municipalities regions hospitals, but also companies like AstraZeneca. So we can work together and benefit the many, as well. And also a lot of the techniques are across industry. Interesting like image segmentation. There's very interesting from both satellites, agriculture perspective, but equally from a cancer detection perspective. There's many parts of the workflow that can benefit you no matter what industry you are in.
George: So Kye, it feels sometimes some citizens are still wary about AI. And do you think this is part of the " every time something new is introduced, they're afraid of it?". I read articles when they were talking about when cars first came out and how society was like, oh, this will lead to the destruction of families as a concept. And then, the telephone came in, and it was the destruction of civil conversations. And then with computers, everybody said, we're not going to be able to memorize things on our own ever again. And things like that. Do you feel there's people being reticent of adopting AI and having AI as part of their lifestyle, and having this part of society just because they're afraid of the unknown?
Kye: Now, I think the general public or the everyday citizen is pretty enthusiastic about new technology and change in general. But I think media are very prone to pick up on scare messages. I would say four or five years ago, whenever I spoke to anyone, I didn't even, I didn't meet anyone who was fearful about AI. Most people were enthusiastic about self-driving cars or bots, helping them in their everyday life.
But after four years, terminator articles, people are picking up on that fear obviously, and discussing that more and more. Yeah, but I would say in general people, if you give them tools that can help their lives and involve them in the discussion, instead of keeping them outside of discussion and saying that there's afraid, they're going to be happy and are willing to change if you really give them the tools. No I don't think so. I think most people are willing to change.
George: So what's happening with AI ethics? Is there some sort of a worldwide consortium that's coming together to decide what these principles should be, or is it still at a country level?
Kye: We set up the AI ethics lab for AI Sweden. And we spent half a year on that process. We spoke to a lot of philosophers and experts all around the country and even outside the country to try to design the process. And the first thing is we didn't come up with one single case of ethics and ethics problematics that were specific to AI only. Everything fell in under normal, like normal ethics cases, existing policies, existing regulations, just at higher pace or maybe a faster speed. So that's the first thing. And then I think the second thing is what we often forget this is really ethical that we are stalling technical development to discuss AI ethics.
When children are falling out of school because school is 150 years old, or people are dying in hospital or getting sick. What are we worried about? I think we need to change the theoretical discussion and start discussing instead what kind of future do we want to design? What kind of future would we want to see 20 years from now?
And what incentives on one side and what regulations on the other side do we need to get there? Because we're very fast, like painting up this scary picture of AI that we're very hesitant to even mention the fact that we are unethical right now. We are discriminating. Whenever anyone applies for a bank loan, we're discriminating. I know and so on. And what AI and data can do is help us be more objective and more transparent.
Diana: I love it. I saw it on LinkedIn, one of your quotes that I really liked, and I wanted to talk about it. You said, "AI can help us stop wasting energy on conflict, worries, fear or politics". Can you go a little bit deeper into what that means to you?
Kye: It's a multifaceted topic. We say, if you range, like from war to just a heated discussion in society between different filings in the media or politics, it's very influenced by power safeguarding and opinions. The same way, if I tell you, an opinion about you or your health, you're going to be threatened. or you are going to feel like it's an attack, but if we look at pure data, then we can collaborate around it. So if we can have AI to cut through the opinions that China is bad, or the US is bad, depending on what's which side you come from. And instead look at the data in the future, then we can have more constructive discussions and less conflict.
Because the most effective method there is to gather people and as a group is to create a common enemy outside you.
George: It's disturbing in a way that the common enemy is not yet, bad healthcare, bad education. It seems like there's other focuses out there that are not as important as these, so yeah, definitely. I think governments should get their priorities straight. We should help them get those priorities straight and AI could help us.
Kye: Yeah, definitely. I think it would be beneficial to have a Chrome plugin or extension that would always rate the information you read from a, both from a scientific perspective, but also from a truth perspective, how biased is this? Or just give you a sentiment score. Is this written to try to influence you or is it an objective fact based.
George: I heard that there are companies, individuals working on this already. I forgot his name, right now. Diana, we actually had him on our show last year, but apparently he was working on some sort of a score to... you would provide the article's source and then the the AI would be able to scroll through and determine the type of language that's being used. And see if that's trying to influence you or that's trying to play with your emotions, or if it's just a little bit more objective in the way that it's telling that story. Yeah. So we have a question here from Kate, and she's wondering which industries or sectors you think will not be impacted by.
Kye: None. I don't think there's going to be any industry or sector impacted by AI.
George: Nice. And we have another one from Baharat and a they're wondering "Are there use cases for AI, for developing sustainable pricing models for producers and participants across the agriculture value chain?". It's a very specific.
Kye: Yeah, I think the technology exists definitely to do that. I think that there's it's a regulatory or policy question in Sweden, we have comparison prices. I don't know when that law was made probably in the seventies. You can always see how much does this cost per kilos. So you, the consumer gets to a sense. And can't be fooled. I think it would make a lot of sense to regulate that just to always share, force sharing of data of carbon equivalent emissions throughout the entire agricultural chain, because currently countries only measure what's happening inside their own borders, not their foreign production.
George: Thank you. Thank you so much for that. Diana, I cut you off a couple of times.
Diana: Yes, one of my last questions. So we talked about all the positive aspects of AI and I was wondering what, if you see any risks with AI and how are they currently mitigated?
Kye: The biggest risk with AI is that it's an argumentation of us as humans. And if we, as humans are inherently good for this planet, then it's going to be all fine and dandy. If there's anything bad in humans, then it's going to augment that as well. So I would say it's very much up to us humans to use this technology the way we want.
George: I love that answer. Talking to you, I get a lot of positivity and a lot of hope for what the future holds and what AI could bring us. And I hope to hear more from you and from more people that think like you, and I'm hoping that with this kind of thinking, we'll just be able to make it happen sooner rather than later.
Kye: Thank you very much. Yeah, me too. Yeah. I'm very excited by 2022, it's going to be a lot of fun, fun things happening.
Diana: So maybe we should book already one December 31st, another show the first and the last show of the year,
George: To have a review of 2022.
Diana: To see what happened in AI 2022.
George: All right. Thank you so much, Kye for being on the lights and data show and talking to us about why society should embrace AI.
If you'd like to learn more about AI, please follow Kye Andersson and definitely check out AI for Sweden.
Diana: Thank you very much, everyone for being here and for your lovely questions. I interrupted Kye, did you want to say anything?
Kye: Thank you very much. It has been a fun conversation and I think keep this conversation up and don't forget to have the courage to discuss and question the big questions in society. I think it's needed for us to be able to really use the potential of AI pull out.
George: Thank you. Bye everyone.