Top professions impacted by AI and skills for the future

 
Curiosities  
AI Curious Episode 3 
AI and the future of work with  
Professor Sarath Kodagoda 
 
Duration 14min 44sec 
 
00:00:00:00 - 00:00:23:21 
 
Hello curious people. I'm Professor Sarath Kodagoda, Director of the UTS Robotics Institute, and I am here to answer your curious questions about artificial intelligence or AI. This is AI curious. 
 
00:00:23:23 - 00:00:56:02 
 
Okay, let's get on to it. Let's look at what our social media community is asking. What does the future of work looks like? Well, you can think about currently what is happening when comparing to what happens maybe ten years back. The work that you do more and more repetitively is more and more replaced by AI and automation. With, for example, if you take the industrial point of view, there are a lot of repetitive tasks. 
 
00:00:56:04 - 00:01:24:14 
 
There can be unsafe tasks. So these tasks could be easily automated so that you won't be doing that. But there could be a point where you will be supervising how it could be done. But I think because AI is everywhere, when you look at even your phone, when you look at Google search, when you look at day to day activities, when you look at social media, AI is everywhere. 
 
00:01:24:16 - 00:01:51:04 
 
So it is influencing the way that you do things not only at the workplace, in your personal life as well. Is AI coming to steal our jobs? Is there any real threat that robots will take over? If you look at what we have gone through as human in 19th century, we have this industrial revolution. Steam engine was invented. 
 
00:01:51:06 - 00:02:20:20 
 
Then we had these factories, factory lines. What happened there? People lost jobs. Unfortunately, people lost jobs. But what also happened is it opens up so much of other avenues relating to that. And then when it comes to 1980s, if you look at how the internet was born, people thought, we are gone, I am pretty sure we lost so many jobs over that. 
 
00:02:20:22 - 00:02:54:11 
 
But reality is now imagine what you are doing with internet. It looks like you can't do anything without the internet these days. So it has come a long way. Now we are talking about as the third generation, AI, robotics, data super highways. Are they going to take our jobs? Well, probably yes, some jobs, depending on what AI can do, what the robotics can do. 
 
00:02:54:13 - 00:03:24:06 
 
But the good thing is that if you look at now, even ChatGPT or any other AI based tools, it has been used in many industries. Unfortunately, there will be some job losses, but it will create more opportunities if you learn how to use it, explore it, be aware it and make yourself comfortable with AI and robotics if possible at all, I think that is going to be the future. 
 
00:03:24:08 - 00:03:48:05 
 
What skills will be sought after an AI world? Well, I can say that AI can't do everything, it can do many things. But if you look at a situation where the human touch is required, for example, if you go to a doctor, I don't want to go and talk to an AI bot. 
 
00:03:48:07 - 00:04:19:23 
 
So there are skill sets which are related to, let's say, emotional intelligence. AI can do so many things related to analysing data, helping out with safety concerns with robotics. But I think when it comes to working with people, it becomes especially for the emotional aspects of it. It's going to be really hard. What can I do to upskill, to prepare for the future of work? 
 
00:04:20:00 - 00:04:47:15 
 
The world is changing in a way that we can't imagine. Every day a new thing is coming. Therefore, I think upskilling, learning new things, what we call is lifelong learning. There are different ways you can achieve some of those. AI or robotics related skill sets, many online courses available out there so that you can upskill at the same time. 
 
00:04:47:17 - 00:05:25:21 
 
If you are very curious, choosing the university courses, choosing the TAFE related courses and there are some other areas we call micro-credentials, meaning that you don't really commit yourself to do a whole degree, but you will pick courses or subjects so that you can upskill the areas that you would really would like to upskill. So in that way, your time commitment, for example, if you are working, you don't have much time for a degree, but you can do a 1 or 2 subjects. 
 
00:05:25:23 - 00:05:54:10 
 
 
 
So in that way you can really bring your skill sets up with your curiosity so that you can be much prepared for what is coming in the future. I want to get into AI development. However, I don't know where to start. I have the theoretical knowledge, but I don't know how to actually apply it. Can you break it down with an example project? 
 
00:05:54:12 - 00:06:22:01 
 
In a very broad sense, first thing you got to know what the problem is. Once you identify what the problem is, second thing is what data that you are going to use and where this data are coming from. Is it coming, is it already available on the web somewhere you can download, or is it that you have to capture the data by yourself? 
 
00:06:22:03 - 00:06:50:05 
 
Next thing is, once you know that data, you get to know whether that data has a label associated with it. To say that what this image tells you. Then you separate that data into two portions. One is called the testing dataset. The other one is called the training dataset. Once you have all this in place, you go to think about what sort of algorithms that you are going to use. 
 
00:06:50:07 - 00:07:19:02 
 
Once you know which algorithm to use, because as you can imagine, there are so many algorithms out there. What you can do is start applying that training or start training that data set in such a way that you are testing, or when you are testing it, your performance is better. So in other words, you are trying to train that AI algorithm to achieve what you are expecting. 
 
00:07:19:06 - 00:07:48:14 
 
And then once it is done, and once you are satisfied with the performance, you can apply to any other data sets that you will have or going to capture. So that's kind of a process of the development cycle that in AI generally adapt. This is exactly the same process that we follow at UTS's Robotics Institute to do our different projects with different industry partners. 
 
00:07:48:16 - 00:08:27:06 
 
One of them is working with Sydney Water to identify defects in sewer pipes or water pipes. What provisions do people want to interact with humans versus robots? There are interactions that we value. Human nature, for example, we don't want to go to a computer to ask, talk about what I am going though. So when it comes to the health care sector, there is a potentially high chance that people would trust people want to go to a real person. 
 
00:08:27:12 - 00:09:03:18 
 
 
 
We normally go to GP's, doctors counselling services, maybe physiotherapy. You don't want a robot to press here and there. There are other situations where this becomes more important as well. For example, teachers, they do a tremendous job in not only teaching the content but mentoring, guiding you, talking about the futures and most importantly, building trust is very important in that sense, 
 
00:09:03:20 - 00:09:41:07 
 
for me, to go to a real person rather than going into a chatbot. So when it comes to emotional attachments, when it comes to trustworthiness, the most probably and human communication interactions people will choose, or people are more intended to choose humans over robots. So what questions does I struggle with? I think it's good to understand. It's important to understand that as a because sometimes we used to think that, well, AI can do everything. 
 
00:09:41:13 - 00:10:17:04 
 
All these problems are solved. But no, there are things that AI is really struggling with. One of them is really about general knowledge or how we interact. For example, think about there is humour that you want to bring into the conversation. I'm not sure how AI is going to respond there. Think about empathy. Think about emotional intelligence that is required when dealing with some of those issues or some of these conversations. 
 
00:10:17:06 - 00:10:50:16 
 
AI is going to be struggling with those. Think about understanding somebody is situation in terms of in a humane way and maybe, helping that person. So when it comes to emotions, when it comes to intelligence in a deeper way, when it comes to ethics, AI is at the moment it is still struggling, but I am not sure in 50 years time what it going to be. 
 
00:10:50:22 - 00:11:20:13 
 
But I am really sure that at the moment, as it is, those are the areas that really AI is struggling with. What role does AI play in addressing global challenges such as climate change? I believe that there are so many things happening in the areas of climate change, and AI is supporting, trying to achieve other goals in terms of climate change. 
 
00:11:20:15 - 00:12:02:21 
 
I can give you a few examples. One of them is AI can be used to optimise our resource use. For example, we are talking about smart buildings, smart cities. There are so many aspects that we, even without knowing AI, is contributing. But in the very high, broader sense, AI is using enormous data that has been collected over the periods for climate change modelling, predicting carbon in order to better predict, better manage climate change related issues. 
 
 
00:12:02:23 - 00:12:39:16 
 
So definitely there is a huge role AI place in terms of climate change. With so much of automation and AI becoming a trend in coursework, like how would you apply in a positive manner to coursework and allow students to apply those learnings from AI to industry? I would like to answer that based on my experience and how I would think whether you can use ChatGPT sort of structures in preparation of this courseworks or writing reports. 
 
00:12:39:18 - 00:13:14:15 
 
My answer to that is I have a certain view on even, you know, using the technology. For example, when you look at a calculator, it is a good thing. You can do the calculations, but if you give that calculator to a, student who is starting, let's say year one, that might not be the best because you need to first develop that addition, subtraction, multiplication skills before you start using the calculator, which does it for you. 
 
00:13:14:17 - 00:13:42:24 
 
Similarly, in this case, the AI may be able to generate reports for you, but I think you also got to know that there are certain skill set that you need to develop in order to, progress on your education. And the other thing is, there could be some research projects that you can bring into the class. We do, we do bring research projects. 
 
00:13:43:01 - 00:14:18:05 
 
We do bring industry projects into the class so that students are really working on real issues or potentially future issues, which is a great insight for them to learn about, to know about what is going to happen, when they are going to join the work. So I think there are so many potentials in integrating AI into the classroom, into the courseworks, 
 
00:14:18:07 - 00:14:44:20 
 
but we got to be very careful not to overuse it, but to use it for the advantage of all of the learning. That was all the questions for today. I hope you learned something new and until next time, stay curious.