So today were going to be discussing skills for the future. And were going to about the evolution of the types skills that are required jobs over time and so today were going to be discussing wefor the future, an were going to talk about the evolution of the type of skills that are required for jobs over time. And what we expect in the future. So how many of you have heard about robots taking over our jobs . How many of you are actually afraid of that . Can i see a bigger show of hands . How many of you have heard about it on the news all the time . That youre about to not have a job, because theyre taking over . All of you . Seriously. And how many of you are afraid of that, that it may actually happen. Okay, about three quarters of you, maybe, kind of, sort of, a little bit. Lets start out by showing the graph. That is a graph of Employment Level in the United States from january of 1948 to 2023. When we take a look at the long term picture of the level in the United States, you can see that the Employment Level in the United States has always been increasing. Yes, its true our population has been increasing as well, but sow have jobs and the Employment Level in the United States. These shaded areas represent recessions. And as you are aware, when were in a recession, we see a decrease in the unemployment rate. So its important for us to stay focused on whats going on in the long run, as opposed to these short term disruptions in our economy. So hopefully this helps you feel a little better about that, but the fact that we have seen an increase in Employment Levels over time, and we do expect to see an Employment Level increase over time as well. When you think about all the changes that we have experienced in our economy from the 1950s, to the 1960s, this is about when computers were being started to be used in businesses and in communications. In the 1970s and the 1980s, heres when people started relying on personal computers. In the 1990s, we started relying on the complex mobile phones we have. And now you can see the number of people applied in our country increasing. So we should not think about computers and robots as substitutes for human skills. Rather, we should think about them as compliments to human skills. And this graph does show us what has happened to the percent of employment in the Manufacturing Sector. As you can see over here, we have seen a decrease in jobs in the Manufacturing Sector. We take a look between january of 1970, to january 2012, you notice how about a quarter, or slightly more than a quarter of our labor force was in the Manufacturing Sector in the early 1970s, and weve seen a steep significant decrease in the employment in the Manufacturing Sector to about 10 or so by the end of this period in 2012. So while employment has increased, there has been a shift in the types of jobs that people are working in our economy today. As you can tell right here the Manufacturing Sector has been shrinking. And this is where a lot of these minimum class jobs are. Many of these jobs are in the midwest, like michigan, perhaps. And you know, in the midwest we have seen the increase in the employment within the Manufacturing Sector, but weve seen increases in employment in other sectors. So in the next graph im going to show you what has happened to employment in selected sectors. So i got this graph from the st. Louis federal website, where its pretty easy to go out and grab a lot of data from the bureau of labor statistics, for example. And i picked some specific sectors in our economy, such as construction, legion, hospitality, retail trade, professional and business services. Financial activities, climate, education, and Health Services. Transportation, warehousing and manufacturing. And in this graph, i went back all the way to 1940 to today, the latest data that i could find in march, 2023. So this top line over here shows us all employees in the Manufacturing Sector. So what was going on in the Manufacturing Sector was until up to about the mid1940s or so, we saw an increase in the Manufacturing Sector, after which, it stayed pretty much stable, pretty much steady, and then we have started to see a drop or decrease in the more recent decades. But were seeing jobs in other sectors opening up. For example, this right here, this slide right over here. Youll see a steady growth. Jewel see a steady growth in jobs. What sector is this . Yeah. Private education. Private education and Health Services. And theres a few others as well. So because of interest in the shift that has been happening more recently, i decided to just go break off this graph, just so we could take a closer look at whats happening recently, and start this graph from january 1970, and where we were ending in the previous graph in march of 2023. So you can continue to see the trend of a decrease in employment of the Manufacturing Sector. We do see a deep decrease during recessions. The shaded areas in this graph are recessions, such as during the 2020 covid recession. But these are short term decreases and eventually, we go back up to our old levels. So when you look at these graphs, i want you to focus on whats been going on in the long run. Whats been going on over a ten year period, a 20 year period, decades, or even over a century. As billy mentioned over there, the private education and Health Service sector has been growing. What are some of the other sectors that have been growing . Yes . Professional and business services. Professional and business services. Yes, i think so, yes. So this has been growing. And which one is this . It seems like its been growing as well. Oh, maybe not. Yeah, i think thats been a long term trend of growth. So which one is this one . Yes . I think its leisure and hospitality. Leisure and hospitality. Its really interesting to take a look at what happened during the covid shutdown when a lot of businesses were shutdown. We saw a lot of a drop in the leisure and hospitality industry. But in the long term, we are seeing a long term increase in that as well. And over here, that are pretty much steady. Maybe theyve been growing a little bit, but theyre pretty much steady. So what are these sectors here . Yes . Construction, transportation and warehousing financial activities. Construction, transportation, and warehouses financial activities seem pretty steady. It is important to see that some factors, such as the Manufacturing Sector has been shrinking and decreasing, that does not mean that there are not Jobs Available in other sectors, because you do see this shift thats taking place in our economy with the type of jobs that are available in our economy. So whats really going on in our economy over time, over the last few decades has been structural changes in the labor market. These early structural long term changes in our labor market. There have been different types of jobs that are being created. Were seeing growth in certain industries, but were also seeing other industries shrinking. And these rapidly changing labor markets are due to two things. Technological progress, and heres where, you know, were all concerned about computers taking over jobs. But i want you to focus on the types of jobs in the Manufacturing Sector. What are the types of jobs in that sector that are so easy for us to, for example, to have a computer take over . Or to program a computer to do . Yes. Like assembly or organizational tasks. Assembly or organizational tasks, why . Why do we think those jobs are easier to get computers to do . Theyre very routine. So you can get a computer to like do that pretty easily. Perfect. Theyre routine. So jobs that are really routine, you can easily Program Computers to do. You give them specific rules. Do this, then you do that, then you do that, and you do that. And its very easy for a computer to take over. Now its these types of jobs that can also be easily outsourced to cheaper waged countries, because these are the types of jobs which we can provide specific rules to workers in those countries, and they can follow those very routine rules that theyre going to be following, and complete the tasks. So given that these jobs are disappearing in our economy, the jobs in the past in our economy or going to be very difficult from the types of jobs in the future. And since these jobs are going to be different, we are going to require different skills. A different set of skills from humans in todays labor market, as well as in future labor markets. Were seeing some labor market polarization, where the types of jobs were finding cannot be outsourced are either the very low skilled jobs or the very high skilled jobs. And many of these middle school jobs have been automated, or sent abroad. And we expect this pattern to continue in the near future. What we need today is we need skills that compliment technology, or skills that computers cannot substitute. So lets think about what computers can do, or computers cannot do. Have any of you tried that robot vacuum cleaner . What do you think about it . Are i just like the convenience. Its very hands off, which i think is good if youre busy. Yeah, its really convenient, its really fast, its very helpful if were busy. Thats why i love it. I bought one last year. Im a busy person. Im just going to turn on the robot vacuum cleaner and it swallowed up my phone charger. Completely swallowed it up, it kept going and going. Lets think about how a human would have been different. Once you see the floor charger, once you see that something is happening with it, and the vacuum cleaner is swallowing up your phone charger, you stop. You pull out the wire. Youre pretty much going to find a solution to a situation that is unfamiliar to you that youve never been in before . Thats something that robots cannot do. Because robots are being programmed based on information thats available. So lets think about what computers and rebots can do, and what they cannot do. So computers, as you mentioned speed, and accuracy. I make mistakes sometimes when im tired. Sometimes im sitting around, and i cant figure out what 2 times 2 is, because im tired. But the computer can be going all night, and its going to be accurate. But heres the weakness of the computer. They can only process information given and they make inforrences based on commands. There needs to be some kind of information given to them. Then based on that information, its going to follow a specific command or a specific rule on what to do next. So the computer executes rules. The computers are programmed with a series of rules that specify an action. You have to tell a computer what to do, and there needs to be Information Available for the computer to be able to execute rules. Humans, on the other hand, we process information cognitively, then we decide what to do next. Our brains are very flexible, and were able to adapt to new situations. We can also draw upon our past experiences, and then use past information, and modify that information based on the situation that were currently facing, and come up with what might be the best solution for it. So a lot of the information that we are going to be talking about in the next few slides come from a paper titled dancing with robots, human skills, computerized work. In their work, they discuss how the types of jobs and the types of skills that are required in our economy have changed over time. Lets go back to discussing computers. What computers can, and cannot do. Computers can substitute for a human in performing a particular task when two conditions are satisfied. The first condition is information condition. The computer needs to have some preexisting information. Computer needs to have the information necessary to carry out a task. The second condition is the processing condition. That information has to be processed in a specific set of rules for the computer to be able to execute the information. The rules can be conductive, or inductive. The deductive rules are the ones were all familiar with. This is a rulesbased logic. Where its a logical stepby step instruction thats programmed into a computer. This is pretty easy to computerize and to outsource. A good example. Can you think about an example where were just programming the computer stepbystep to follow a specific set of rules . Yeah. Like the way you Program Thermostats in a house, maybe . So if you want to keep a temperature above or below a certain threshold. Yeah, so youre telling the computer i want it at 70 degrees. I want my home at 70 degrees. So if the temperature in the house is above 70 degrees, thats a very good example, thank you. If the temperature in the house is above 70 degrees, then the ac turns on. If its below 70 degrees, the ac does not turn on. In that, were giving the computer specific rule by rule, stepbystep procedure to follow. But rules could also be inductive. Inductive rules would be something that pattern recognition, where the Information Processing cant be articulated as a series of logical steps. With pattern recognition, we could use statistical modeling to look for patterns in data. And based on these patterns of data in the past, we can then try to predict a solution for something that has not happened before. When you think about ai, and all you know, writing the essays and so on, this teacher is concerned about plugging in a command. I actually tried that out. I got a bunch of junk. Try it out, one day. But what the computer does is it digs out information that is already in the internet on the specific topic. And then it processes this information and it spits out something new based on what it thinks we want. Its not always correct. Thats a little harder to program, but thats someone where they were trying to use information from the past, process that information, and come up with something brand new, or what i like to call, maybe, new output, perhaps. But when i tried it out understand i found that it was not able to process those words into that ability to get to that point just yet. So thats something that computers are unable to do. And inductive rules are a little bit harder for us to Program Computers to do. Some of it we have figured out how to do. But theres still a lot that computers at this time or in the future might not be able to do. Heres some uniquely human brain strengths. Fflexibility. We have the ility to process and integrate many kinds of information to perform a complex task. We can solve problems which theres no standard operating procedures. We can work with new information, if you already do the same thing and i ask you to write an essay on education, your essay would be very different. You can go online and use the internet as a compliment to your learning, and you would process all of that information at a deeper level. You would make sense of it, and you would communicate it affectively in a way that answers the prompt that i wanted. And dont forget, we have common sense. We know when to stop. When your vacuum cleaner is swallowing up the phone charger. These are some things that computers do not have yet. And this is a table that i also did borrow from the work my level. Theres rules based logic, as we mentioned before. Those productive rules. And with deductive rules, its pretty easy to program a computer to do. Anyone use turbotax . Its pretty convenient. Step one, whats your income . Step two, your beholdings, step three, your savings. Enter in the information, no, move onto the next step. That will be an example. Or issuing a boarding pass. In the past, you would have to go up to the Airline Attendant lineup, give your passport, or your identify to the attend he want, and the attendant will check into the computer and issue a boarding pass. Today, you can just get your credit card, stick it in one of those computers, and the computer matches your name to their system and issues a boarding pass. If the computer is unable to match your name to the system, thane then then youll receive a message, go see the attendant. These are the types of jobs that are easily replaced by computer, and will continue to be replaced by computers. At we move out here, increasingly difficult to program will be pattern recognition. It is possible for us to Program Computers up to a certain point with these inductive rules. So each recognition on theory. Everyone has a accent. Lets say that were asking siri a question. Siri is programmed. We can program siri to recognize these words that are pronounced slightly differently across different people. But it is not as easy as the deductive rules that we are programming to computers. Predicting a mortgage default rate. With predicting a mortgage default rate, if a bank would like to predict whether a customer would default on a mortgage, what they would do is they would get information about the customer, such as, you know, maybe the customers credit score, how much the customer owes people. The access that the customer has, and they will compare it based on their data set of previous customers. They might have a data set of thousands of customers in the past and take a look a the likelihood of a customer with a similar profile, defaulting on the loan. So thats a little bit more information process that takes place, were able to Program Computers to do that. But theres certain things, at least now, that remains human work. Rules that cannot be articulated. And if the information cannot be contained. For example, writing a legal brief. Lets say that there is somebody that you just suspect committed a crime. Amend you suspect that this person broke into a house for hats at midnight. Ran in, and i dont know, stole some food at midnight or Something Like that. Its your job to go request out and try to find that information, to connect all the pieces together.