You and I chatted before about the need for a license or credential being a barrier to entry into professional fields. Of course. An external license also ensures that the license holder has the minimum knowledge for that field, to protect the public, but the idea that people can just hop from one career to the other isn’t going to work easily because of barriers to entry. People in these industries keep upping the ante to keep people out of the profession. The less people in the profession the higher fees can be charged so the consumer ends up being HURT not HELPED.
@prof2dad Linked the wrong article! Sorry!
Here is what he said:
http://www.businessinsider.com/mark-cuban-liberal-arts-is-the-future-2017-2
@PurpleTitan Right now, Royal Caribbean has a bionic bar that makes drinks. Manufacturing jobs will always be necessary. Whether or not those jobs will be in the US will be debatable.
I can only speak for myself though when I say, I would prefer robotic made food with high quality ingredients (can be done easily for firms with a bionic bar) over human made dishes with average ingredients. The real business will be for those who would be making the goods that the robot will use. Unless, of course, there is a robot doing that too. 
You mean we have to start this discussion all over? 
@saillakeerie I don’t think so lol, I think the gist is that jobs involving math will soon be obsolete as more and more things become automated. I think Mark is right. If someone can truly master AI and figure out all of the nuances, there will be an app, robot, or software made for literally everything. It could possibly be able to function on it’s own.
@i012575 Fewer humanities grads = more stem grads = fewer available stem jobs in the long run = more demand for humanities in the long run?
Math has been mastered and automated long ago. What financial people do is interpret the math.
Robo advisors are gaining market share, especially with young people, in the Fin Services industry. I hardly think finance is immune from losing jobs to computers/automation.
Very few people have jobs that require truly high levels of creativity. Artists need high levels of creativity. Academic researchers need high levels of creativity. Most white collar jobs require what I would term creative problem solving, thinking of multiple hypotheses, evaluating those hypotheses against the evidence, formulating a list of solutions, and running a cost-benefit analysis against all the options. True AI will be increasingly good at creative problem solving.
@CaliCash No problem.
The finance industry is of course not immune from automation. Those jobs spending a lot of time on data processing could be replaced by automation.
For wall street banks, as I mentioned in my earlier posts, automation is currently undergoing largely in the middle and back offices where IT, HR, repetitive administration, compliance jobs are been streamlined. But for finance/economics majors who tend to occupy the front office, their likelihood is actually largely residing on their soft skills: relationship, networking, and interacting with clients. They would need to have good and clear financial/economic intuitions and need to be able to articulate and communicate with clients. These tasks and functions are generally very difficult to be automated; no institutional clients would pay high fees to talk to a machine.
Relationship is so important in banking that I remember 20 years ago when I was a doctoral student, my banking seminar was entirely focusing on the issue of relationship in banking. Even though my professor used to be a FDIC auditor, he said nothing about regulation during that semester. Now looking back, I wish he could spare 1 or 2 meetings on sensible regulations.
Tons of financial market trading has been/is getting automated. For all that work that used to be done by human traders, one doesn’t need employees who can “need to be able to articulate and communicate”.
“no institutional clients would pay high fees to talk to a machine.”
Institutional clients are automating as well. In competing for business, lowering expense ratios is key and automation allows that, so no need to talk to your institutional sales rep if your trading is increasingly automated.
The financial services industry has seen rapid automation for the past few decades and it will continue.
@doschicos You are talking about trading and asset management (buy side), whereas I was talking about banking (sell side). The reason I focused my discussion on banking was because on CC many users equate finance with wall street banks.
Yes, some trading, asset management, and brokerage/advising were automated “long” time ago; e.g., Scottrade, algorithm trading, etc. I did not talk about it because I thought by now it is evident and my acknowledgment, “the finance industry is of course not immune from automation,” should be sufficient. After all, in every decent-size town/city, we are likely to see at least one Scottrade office, right?
You can google automation and investment banking and you’ll find plenty of hits that that area won’t be spared either. Relationships were valued in the past but efficiencies will make that less important and the front runners to automation are often the winners. Bottom line: automation will equal $$ savings and that will win out over relationships.
Many banks today are megabanks. That is, they have many business divisions, ranging from investment banking division (IBD) all the way to depository banking or even brokerage. When you google investment banking or the name of a megabank, those reporters tend to lump the automation of trading, retail banking, and brokerage with “investment banking” in IBD together.
I think it is worthwhile to think why there is a division in an investment bank called investment banking division (IBD). For those who do not know finance, they may think the term is quite redundant. But it is what it is for a reason.
Relationship is fundamentally important for banking. I hate to go to the theory. But this is the main idea: banking has a moral hazard problem. Relationship endows access for monitoring that can be used to address moral hazard. If moral hazard is mitigated, value can be created for both banks and clients. The following paper was published in the Journal of Finance (2012), the most influential finance journal, that again demonstrated banking is about relationship (we have tons of papers showing similar things in the past 50 years): http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.622.211&rep=rep1&type=pdf
Nope, I’m talking about investment banking and do know what it is, but thanks for mansplaining. I disagree with you but we’ll find out down the road I guess.
@doschicos I respect that you disagree.
I believe that I have been conducting myself well. The use of the word, “mansplaining,” is unjust.
@prof2dad Well, how about “acting professorial”? Your assumption that I am clueless about the banking world, both IB and otherwise, I find a bit off-putting. Reread your comments with the assumption that others DO know the industry, and perhaps you’ll see what I’m talking about. Besides, trading is both buy and sell side, is it not?
I agree with @prof2dad Investment Banking is about relationships and analysis. Of course the industry has and will continue to automate (I can’t think of a white collar industry that has automated more than Finance/Accounting) - but big data needs interpretation by humans. And big institutional investors need someone to buy them a steak dinner
There’s a lot of grunt and analytical work in IB that can and will be increasingly automated. It’s not all about dinners at Peter Luger and signing the deal. Like much to do with automation and job prospects, the higher ups will often do fine but down the pipeline, there is plenty that could see continued automation through AI.
That’s so far away that you might as well talk about warp drives and photon torpedoes at the same time. All of the AI systems I’ve seen in action are strictly mechanical (at the fundamental level). While they may tolerate inputs outside of their “training”, they have absolutely no idea what’s going on. For example, I’ve been working on Landsat 8 imagery lately. In a nanosecond I can look at an entire image and determine what’s a cloud, what’s a cloud shadow, what’s a hill shadow, what’s a city, etc. Now try to use computers to do the same — boom, cities are mistaken for clouds, forests for water, etc. There are thousands of papers on this topic alone. None of the algorithms work reliably because computers simply have no idea what they’re “looking” at. So I bring in all sorts of non-visible imagery data (infrared, etc) to try to help in the process. Still, chunks of cities are mistaken for clouds, bright beaches are mistaken for clouds, etc. A four year old kid wouldn’t make those mistakes. And this is with a fairly well-behaved set of input data. The smarter the algorithm, the dumber the output!
What do the companies specializing in visible satellite imagery do? They use brute force. They grab 20, 50, or more images of the same Landsat scene and for each pixel sort the 50 samples by brightness then average together the dimmest ones (there are many variations on this same theme). Even with that, clouds can still get through in areas of persistent cloudiness (especially mountainous tropical islands).