We've heard these terms get thrown around a lot and I bet we all have somewhat of an idea of what these are. At very least we use them in our lives occasionally. Consumer products like Google Home, Alexa, Siri, and the little paper clip that used to pop in Microsoft Word can all be considered Artificial Intelligence (AI). These little "personalities" are smart, they help us in our daily lives: run searches for us, order pizza for us, and even ask if we need help. So I think the real question is what is machine learning and how is it different?
Machine learning is basically what it sounds like. It's a machine that can teach itself or at very least learn over time. AI that we are familiar with are pre programmed to do specific tasks and those tasks expand usually through software updates (just like new features on your phone). But machine learning is predominately programmed to a broader focus but has capabilities to teach itself new tasks (not to say that it doesn't also get updates).
For example there is a map game that is tested with these types of programs. Basically, there are objects on a map and and the computer is supposed to find them based on directions given. In this scenario AI would be pre-programmed with what North East South and West are maybe coupled with the ability to run objects through a database to determine what is what. Essentially, they can look up what an object is and then determine the directions to get to the object with 100% accuracy (as long as the object exists in the database). Machine learning on the other hand can start with a blank slate and learn from trial and error, failing at first but then learning what North East South and West are and even asking what objects are through language learned.
So what does this mean for white collar jobs? Bad news folks...but maybe not for a while. For now, middle-men are in trouble. Third party services are at risk for machine learning to take over. This makes sense considering middle-men typically offer a connection more than an actual product or service. Third parties connect 2 groups together in an efficient manner but it's being recognized that AI and/or machine learning can do it more efficiently and the costs to do so are becoming more affordable.
Take Uber for example. Their app shows a map of nearby drivers through GPS and always gives you the option to be picked up by the nearest driver. Uber is currently moving this system to third party logistics. Instead of having a consultant call a truck to ship your goods and collecting commission both from you and the truck driver Uber steps in and gives you a map of the nearest trucks allowing you to easily do what you used to pay someone else to do. Machine learning gets thrown into the mix by linking more than one shipping destination and minimizing distance driven, maximizing cargo delivered, all while providing a cheaper rate to both parties.
Another industry that is being affected right now is financial technology (fintech). It's one thing to ask AI to automatically give you all the ratios you want and then pick stocks based on your analysis and criteria for those ratios, but what if you could ask a broader question? What stock will make me money? Machine learning could find statistics and ratios that link up better than what you originally thought. So not only would it calculate and find all the ratios you are interested about but it could suggest more and even perform the analysis without you based off of what it has learned (decades of news, financial statements, and performance history).
Again, we may be in the heyday of AI, but we are only scratching the surface of machine learning. Right now you don't need to fear for your job, but it might be a good idea to stay ahead of the curve and learn to work with machine learning before it starts replacing jobs. I think right now we are at the equivalent to when machines started automating production factories - which shifted jobs from the floor to repairmen for the machines.
Links:
https://www.linkedin.com/pulse/uber-freight-its-just-truck-drivers-whose-jobs-risk-john-mcdermott
https://letstalkpayments.com/15-players-that-use-machine-learning-in-fintech-space/
Hey there, Fernando! Great Tech Briefing you have here. I actually find the machine learning aspect of your post to be extremely interesting. 1, because of how cool it is for these machines and products to have the ability to actually teach themselves, and 2, because we use them in our everyday lives.
ReplyDeleteWhat I think is so interesting are the pros and cons that come along with all these technological advances that actually make the majority of our lives much easier and quicker. What it seems like to me, are that these advances make daily activities to the general public much faster and easier to complete, but for those that are in higher up, white collar positions actually affect them more negatively due to their career possibly being replaced.
I also see how you mentioned that this may not all be in full effect for quite some time, but it’ll be crazy to see the transformation over the years as we all enter the workforce, and how it may effect each of us in a different way. For myself, being a Finance and an MIS major (as mentioned with the financial technology industry) how that may play into the careers that I have in the future. Hopefully not too negatively.
Just out of curiosity, as you mentioned about staying ahead of the curve before they replace our jobs, are there actual ways to do so? It seems that these advances will happen no matter what, so how do other recommend staying ahead in order to not have our jobs completely replaced? Just curious if that had been mentioned in anything you read.
Fernando! I also really enjoyed your article, Do you think that using AI as well as machine learning will effect the economy as a whole? right now in Fin Tech, they do use a lot of machine learning and algorithms, but that has also been utilized in the past, but the issue here is an algorithm is only as good as its programmer (for now). Do you think we'll see more regulations in regards to machine learning or complex algorithms especially in industries such as finance?
ReplyDeleteI agree that there will be a huge development in machine learning and AI in the near future. Machines can do better job than human in many tasks and they can also help people to perform certain tasks such as a medical surgery. There is always a concern that whether machines will replace human in some field and cause a large loss of jobs for the human workers. My opinion is that machines will replace human in certain fields eventually, but by then, human should be relocated to other jobs. Government regulation and laws can also be used to make the transaction smoothly.
ReplyDeleteI thought it was really interesting that you pointed out that machine learning can replace the routine tasks but are also starting to specialize in other areas as well. As you mentioned that there will be concern for the middle class jobs being replaced. I would think though people would start moving towards careers more concentrated in technology and machinery to stay ahead of the curve.
ReplyDelete