AI is revolutionising all the industries. In many cases, AI is making our lives easier and it's for free us to do something like more of the creativity, more of the thinking, more of the strategy.
[Val] Hello and welcome to the Be Connected podcast. I'm Val Quinn and I'm a technology commentator, broadcaster, publisher and your host of the Be Connected podcast. So, many of today's machines are programmed with artificial intelligence, and that's to think like humans and mimic our actions. So where can we find artificial intelligence or AI in action in our everyday lives? And what are the benefits and risks of the technology? And should we even contemplate a future where AI enables machines that are conscious and self-aware like a human being? These are big questions. So joining us today to explore these ideas is Steve Nouri, a data science leader who's changed the way people look at artificial intelligence. Steve is an author, an academic and he aims to inspire people about the latest technologies. He's recently founded AI for Diversity and that's a nonprofit global initiative with more than 4000 volunteers, and that educates communities about artificial intelligence. Steve's also an influencer on LinkedIn with 500 million views last year, so he's been busy.
Steve, it's really great to have you with us.
[Steve] Thanks, Val. It's pleasure to be here.
[Val] That's great, look, I'm really excited to talk to you about AI, 'cause it's of my favourite subjects and it's really something that has the potential to change the world. So why don't we just jump right in and right off the deep end, can you describe for us what exactly artificial intelligence is?
[Steve] All right, that's a great question. I'm just going to jump into it and give you a simple one that, I think, is the closest to the right definition. A lot of people might disagree, but that's it. I think when machines learn how to deliver tasks like the way a human do, this is called artificial intelligence. It is usually not programmed directly and they learn indirectly from the ways that we deliver these tasks. And also we have machine learning, which is pretty synonymous with artificial intelligence in my opinion and a lot of people that are practitioners agree with this, but in reality, the only ones that are delivering the value are machine learning algorithms and they are widely used in industries and I can tell you that it is probably the 99% use cases of AI.
[Val] So is there any possibility that computers are really thinking, do you believe? Or is that more of just a way that we describe what our brains do? Is there some thinking going on by a computer?
[Val] It doesn't look like they are thinking, to be very honest. I know that there are some controversies around conscious and sentient AI. They are not, at this point that we are talking, the AI is just a mathematical algorithm. So this is pretty much the underlying mechanism of machine learning.
[Val] I see. Well, let's talk about maybe some examples of artificial intelligence that are here and now, being used in everyday life. What do you think are some of the most common ones?
[Steve] I would say that one of the most common and interesting ones is the algorithm that is keeping us watching TikTok clips and videos for a very long time. It is called the recommender system and, essentially, the recommender system is an AI algorithm that understands our watching patterns and engagement patterns and it will come up with a complex understanding of what our brain likes and what makes us interested into watching more and then it'll recommend those contents as much as possible and sometimes, even, it will be adventurous by adding some content that we have never seen, but we might even be interested. So this is a very interesting, complex algorithm. I think TikTok is the one that is using it in the most notorious way possible. Other websites and services are also using recommender systems like Spotify, Netflix. They're also suggesting some videos and some music to listen. All of them try to help by suggesting things. Also, Amazon uses recommender system, which is pretty useful. I sometimes tend to buy some of those recommendations. That's one of the use cases of those algorithm that are very helpful for humanity. In some cases like TikTok, it's a little bit different, because it's directly helping the business. So it's a mechanism to hold your attention, so you just want to watch more and go to the next one. So that's unfortunately not, I guess, the most constructive way to use AI.
[Val] Hmm, I can certainly attest to the seductiveness of these recommendations systems like TikTok and for those of our audience who may not have tried TikTok, it serves short videos to you based on things that it thinks you're interested in. And I thought I would be immune to TikTok and I can tell you that this AI does work. I'm embarrassed to say it. I go onto TikTok and half an hour later, I'm still watching videos and it's pretty incredible what it can do.
[Steve] I guess it's one of the most seductive AI applications in the world. Apart from these, you have it with image processing and facial recognition are some of the very famous and most used applications. When you log into your mobile phone or your laptop, it is facial recognition. It uses your image to understand the facial features and it matches the patterns with the database. Also, even voice recognition, like when you talk to Alexa or some other applications that are available on a mobile phone, we just ask for delivery of some tasks and then the machine learning algorithm and AI would understand our voice. So you can just send an email or you can call someone by just saying, "Call my friend." And that is very handy in some cases.
[Val] And also actually, if you wanna learn a bit more about that, we have on our Be Connected site some really useful courses where you can learn about speech recognition. So check out our 'Introduction to smart homes', that's a great topic on the Be Connected site. So what does the future hold for us in regards to AI? I mean, after what we've just talked about, what's coming, do you think?
[Steve] So for AI, it's very difficult to predict a distant future, but I can just tell you based on the trends, it seems that AI is entering into some very complex realms like creativity and invention. We also have better algorithms, we have better computer power, we are generating terabytes of data. So it seems very clear that AI is going to be a major part of our life.
[Val] So why don't we explore some of the myths about artificial intelligence and look at those risks as well? So for example, will AI have a huge impact on the job markets and basically, the jobs that we used to do won't be there anymore? Is that a risk?
[Steve] I don't think so. So the reality is, yes, AI is automating a lot of jobs, but at the same time, the pace is not what a lot of people outside the industry were expecting. Those myth are not true that everybody's going to be out of job within five years or so, because we are replaced by AI. It seems that AI still is more of an augmentation, helping humans to deliver their jobs better, faster, easier, cheaper. And in many cases, we are evolving those jobs. Yes, at some point, autonomous cars are going to replace a lot of drivers, but then, at the same time, it is evolving the whole industry. You need to deliver different services to those new industries and there would be need for a lot of individuals to start working. So that's a great question, but I don't believe that that's a real threat.
[Val] Well, that's good to know and reassuring. But what about the idea that an AI and super intelligence will just surpass human intelligence and just take over completely? Is that vision of the future possible?
[Steve] I think, like I briefly mentioned, right now, we don't have any evidence or even understanding, first of all, how to get to that same level of human intelligence. Surpassing human intelligence is just something beyond our imagination. It is a possibility and we always are ambitious, but also, we want to be prepared if that happens, so we are talking about it. But as much as we would like to get there and it sounds interesting for a lot of people, right now, it's more of a science fiction than the science itself.
[Val] There's sometimes talk about AI becoming sentient or conscious. I mean, is that capacity possible in a machine, do you think?
[Steve] That's another interesting topic. I had a panel discussion about it a couple of months ago and these are the favourite topics among technicians and geeky people like myself. But essentially, first of all, we need to understand what does each of those terminologies mean, because what is conscious, who is conscious? Who and what does sentient mean? So that is the first step. But then if you just go by a public understanding of those terminologies, it seems that the AI that we have right now, it is not conscious and not sentient. So couple of months ago, I think, it was OpenAI chief scientist who announced that their large AI models might be slightly conscious and I don't know what does that exactly mean? What does slightly mean in this context? It seems that you want to get away with claiming something really huge and then you put just a slightly in there to make it a little bit less aggressive. But then the respond was by a lot of professors and the senior scientists of many different companies that the current AI now is not even slightly conscious and we don't have a mechanism to make the AI conscious right now. So that particular bit that makes it conscious is not, right now, working. And then the sentient AI is another interesting news that came out a couple of weeks ago by a software engineer at Google. And essentially, the reason that a person thought that AI is conscious because he was talking to a conversational AI and for everyone listening, conversational AI, it's not a neutral, unbiased language model that will give you the truth. It's a conversational AI, so it needs to continue the conversation. And when the software engineer asks the AI that, "Are you sentient?" Or "Do you have feelings?" Or some questions like this, based on my understanding, that particular mechanism came into the play and made the AI to say, "Yes, I am sentient," or, "Yes, I have feelings," or "Yes, I am conscious," or whatever. The answers might look very convincing, just because the quality of the answer and because of the sophistication, you would feel that machine is genuine. But it is just the quality of the, I guess, natural language processing and the language models that are in AI algorithm.
[Val] Wow, I guess, so the lesson there is don't ask a conversational AI if they have feelings, because it's designed to answer your questions to keep talking. What are some of the positive ways that AI can be used to help us in the future? For example, like doing repetitive tasks or very detailed tasks? What are the upsides?
[Steve] Yeah, that's true. All being said, the AI is going to help us in many different ways and the technology is more of an enabler than a risk to the humanity. So there are couple of challenges that we talked about, but AI is revolutionising all the industries. Many cases, AI is making our lives easier. Banking industry is using AI for fraud detection, for giving you recommendation about how you manage your funds and your finances. It is also helping auto automobile industries. Cars are being more autonomous. Even if they are not fully autonomous in the short term, they're still going to help you to mitigate some of these risks that it will avoid crash, it would give you suggestions about different directions that might be easier or faster to go. AI is taking those tasks that are not intellectually stimulating for human. They're not interesting. They do not give us any good feelings and it will free us to do something that we are more good at, at least right now. Like more of the creativity, more of the thinking, more of the strategy. Even in future, a doctor doing a surgery or doing diagnosis you would have AI generating suggestions, making these operations cheaper, faster, more accurate and also, they can calculate very complex equations very quickly. So that's very obvious that we are going to get a lot of benefit out of it. And that's where that people like me are very optimistic and happy that we are pushing the usage of AI in many industries these days.
[Val] Well Steve, that has been an absolutely fascinating conversation with you today about artificial intelligence and where it can take us. Really appreciate you spending the time with us today and thanks for all your amazing advice and we'll talk to you again soon, hopefully.
[Steve] Thanks you very much for having me, Val. It was great.
[Val] If you like what you've heard, please subscribe to receive all the latest episodes and leave us a review to help others find us if you're listening on a podcast platform. And remember to visit the show notes for information on anything we've covered here today, including links and other useful information. And for more about today's subject and to discover other great topics, too, go to www.beconnected.esafety.gov.au. That's www.beconnected.esafety.gov.au. I'm Val Quinn and I look forward to your company next time.
Be Connected is an Australian government initiative developed by the Department of Social Services, the eSafety Commissioner and Good Things Foundation Australia. Be Connected builds the digital skills, confidence and online safety of all Australians with engaging online learning resources and a network of over 3,500 community organisations to support them to thrive in a digital world.