Will robots replace your consulting or financial career?

I was recently reading a couple of books - Rise of the Robot by Martin Ford and Humans Do Not Need Apply by Jerry Kaplan. Both books share insights on the demise of both the industrial and knowledge eras. Both books provide key themes and stark warnings about the future of automation and its potential impact on humanity as we currently know it.

Examples cited in the books:
Artificially intelligent systems double in capacity every two years. AI drives cars, runs restaurants, writes articles, creates original art, composes symphonies and builds houses - and that's just the beginning.

In stock markets, "high-frequency trading" (HFT) systems skim vast profits, risk-free. These rapid HFT programs may fuel market chaos, as in the "Flash Crash" of 2010. Today's "high-frequency trading" (HFT) systems complete transactions tens of thousands of times faster than humans can. In fractions of a second, HFT systems buy the stock low and sell it high, skimming the spread between prices. Some systems analyze market "sentiment," finding subtle, fleeting correlations among diverse stocks. HFT systems may "smooth" the market, but they exacerbate the problem of wealth inequality.

A near-total replacement of the knowledge worker could happen by 2035 in addition to extinction of careers in manufacturing, transportation, and agriculture.

Some experts see opportunity - machines enhancing human intelligence and longevity, for example. Others worry about machines taking over.

What do you think?

Anonymous Monkey, Upload Your Resume - Land a Job

Members that upload a resume get 2.3x the number of interview invites through the Talent Oasis. Learn more.

Comments (169)

Mar 28, 2017

Matter of when, not if

Learn More

2,037 questions across 209 consulting firms. 11 Cases developed by a McKinsey Associate, 10+ hours of video. The WSO Consulting Interview Prep Course has everything you’ll ever need to ace your consulting case interviews. Learn more.

Mar 28, 2017

@Predilection" What is interesting about this concept is the macro economic impact and need for a supply/demand market - if no one has jobs or money to buy "stuff" how will companies make money?

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 2
    • 1
Best Response
Mar 28, 2017

Please make these posts stop

    • 11
Mar 28, 2017

thanks @DeepLearning - I think I got the memo LOL.

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 2
Mar 30, 2017

Lmao 8 SBs....

Mar 28, 2017

We almost made it one day without one of these Terminator esque posts. I also think the OP is a troll.... at least 3 threads today all relatively open ended articles and news sources...

    • 2
Mar 28, 2017

@Wallstreetneversleeps Not a troll, just finished reading a couple of books and wanted to discuss. I guess it is a topic on the minds of many. Thanks for responding and letting me know.

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 2
    • 3
Mar 28, 2017

For the last time, humans and robots are COMPLIMENTS and not SUBSTITUTES. Who do you think programs these computers? Who develops the pinging/gaming strategies that these HFT firms use to skim profits? Who analyzes and responds to the strategies that other HFT firms are using and creates a new algorithm to beat the others?

Computers and Technology systems are nothing without human insight, forethought and innovation.

    • 1
Mar 28, 2017

@LReed I too had the same initial thinking until I read a couple of books that really went into broader detail on the topic and discussed what I initially thought was science fiction. Literally. For example, robots eventually fighting for legal personhood, robots managing people, etc... it's pretty insane concepts. Rise of the Robots is an excellent read.

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 1
Mar 28, 2017

I sure hope robots don't achieve legal personhood. Then every time I drop my iPhone I'll be accused of murder!

    • 5
    • 1
Mar 28, 2017

We are far from this. The technology is not there yet, nor is there money to be made off of it either. We barely got self driving vehicles underway.

Someone has to be there to push the "on" button, and service the machines when they are in maintenance. There are updates need available and people who can design AI framework, test it, etc.

Robots and technology tends to create more issues than solve them :(

    • 1
Mar 29, 2017

@Darkasing_619" According to the 2 books I've read, the technology has been around since the 70s and was purposefully stalled. So it is actually already here.

I can't recall in which book, but one of the books gave an account of AI development discoveries at a computer lab back in the 70s; the programmers saw the potential of this capability and how quickly it could evolve & adapt on its own - so much so that it apparently frightened the programmers. they feared their jobs becoming obsolete, so they purposefully created unstructured coding languages, making computer programs very prescriptive and finite in utility at the time. For example, code was written to solve a math problem or to play chess, but the program couldn't do anything else.

Apparently given the post scarcity era we are entering in (thanks @atleastimnotabanker that's a good word), AI programming has evolved to become much more structural in design. Pandora's box was opened. Structural, genetic programming is now underway to create an almost Darwinism-like environment - computers are now adapting, regenerating new code, throwing out "bad" code all by itself. This is currently happening 24/7/365 at an exponential rate.

I wish I had answers, but I don't. The more it is discussed, the more it will be understood.

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 2
    • 2
Mar 29, 2017

the IBM stuff sounds like complete BS: could you source please?

Mar 29, 2017

AI will inevitably replace your job and if you believe the people who should know best (i.e. the ones that build the AI), this change will happen much earlier than most people expect.

Also your question about "What happens if nobody has job and to whom will companies sell to" misses the point.
The concept of Supply & Demand requires scarcity of resources. With exponential advances in AI, we are quickly entering a new post-scarcity world, so the old concept of making money by selling your skills/products will no longer be valid.

As a result, there will most likely come a short period of great changes very soon.
If AI takes over the low skilled jobs in lets say the next ten years (e.g. bus/taxi/truck drivers, waiters, call center operators, etc. - it's already happening today and the speed will only increase with those systems getting better and cheaper at an exponential rate) there will be millions of unemployed people who simply don't have the skills to get any other job - all jobs that they might be suitable for will also get automated in a matter of months after.
Society in the US (or anywhere else for that matter) is simply not prepared for such an increase in unemployment. If you have let's say a 10-20% increase in unemployment in a short period of time, you can be very sure that no matter what currently thought safe you have, you will be heavily affected.

    • 3
Mar 29, 2017

@atleastimnotabanker Thanks for the additional context and your insights. What are you doing to protect yourself and your family from the inevitable?

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 1
Mar 29, 2017

Meh, if it happens it happens. Not really worth wasting brain power on thinking about the what ifs of the situation given it is 100% out of my (and our) control. The people who develop these systems have top PhDs in top hard quantitative fields, they'll be able to automate pretty much everything eventually. They're not on my playing field so I don't think it's worth wasting energy on.

Mar 29, 2017

@CanadianEnergyBanker Thanks for responding. I think a lot of people share your sentiment, particularly those with less than 15 years left before retirement. It's easy to talk about it on a forum, but it is hard work to think about it and actually start doing something with the information.

If one is planning on working beyond 2030 - whether at a firm or independently - they should think about how to leverage the power of anticipation and time in order to be prepared to maximize on the opportunities that lie ahead.

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 1
Mar 29, 2017

There's going to be either two routes. One where this is all overstated and there's nothing to worry about. Or one where estimates are in line with reality. I think there's no need to prepare in either case because there isn't anything you can do. The same ways that lead to financial independence now are the ones that will lead to independence under AI. It is always a bad idea to have one primary source of income, especially when that is a salary. Not much changes with AI except for how true that statement will ring.

Mar 29, 2017

this is the sort of thing that gets the "bug out" community all wet and excited: if the nightmare scenario of this thing materializes then there will be no need for humans for anything as the robots can do it all themselves. I think we are a long way from that and I also think that we as a species would/should curtail their development before reaching 100% dependency. A computer can never replace the "gut feeling" you need in sales or negotiations which comes from experience and having lived that life. No computer can ever be as sophisticated as the human brain. To that end, why worry about it? even if my job will disappear it's not as if I can just wake up one morning and say hey i think I will be a software engineer now. if i could have done that I would have already so why waste energy in worrying about something I cannot control?

"I'm talking about liquid. Rich enough to have your own jet. Rich enough not to waste time. Fifty, a hundred million dollars, buddy. A player. Or nothing. "
-GG

    • 1
Mar 29, 2017

Artificially intelligent systems double in capacity every two years. AI drives cars, runs restaurants, writes articles, creates original art, composes symphonies and builds houses - and that's just the beginning.

This is where it gets wrong. AI isn't going to follow some Moore's Law kind of BS - it's going to plateau very soon. What will expand are new applications of AI. We'll see automation in lower pay-scale jobs.

That being said, AI cannot be the be all, end all. IMO from my observations of the work of some of the top universities and companies around the world working in deep learning, the kind of world envisioned by most of these engineers is not one in which robots coexist as equals with humanity, but rather as subservient husks without sentience. Yes, everybody agrees that machines should not be given sentience - everyone of them.

That would result in a world one notch below the one depicted in Asimov's I Robot. I'm guessing that Asimov's three laws will then prove to be the bedrock for AI boundaries, owing to their popularity amongst AI researchers.

I can't recall in which book, but one of the books gave an account of AI development discoveries at an IBM lab back in the 70s; the programmers saw the potential of this capability and how quickly it could evolve & adapt on its own - so much so that it apparently frightened the programmers. they feared their jobs becoming obsolete, so they purposefully created unstructured coding languages, making computer programs very prescriptive and finite in utility at the time. For example, code was written to solve a math problem or to play chess, but the program couldn't do anything else.

This is utter BS. AI research wasn't purposely stalled because some righteous researchers wanted to save humanity - it was because we hadn't developed the tech for the insanely high processing power required. But once people decided to utilize GPUs instead of CPUs (this is only one factor, but key), AI research resumed making intense breakthroughs. The difference was that in the 70s and 80s, research produced a huge trove of theoretical knowledge, and now it produces potential applications. OP, you need to revise your reading list - start with Asimov, Penrose and Minsky.

That being said, I'd be happy, not cautious, if a robot were around to help me when I'm old - can't really on the snot-faced progeny these days. I'm also sure that humans will still be at the top of the pyramid, so you can be assured that your consulting career is not going to be replaced by some metal hunk named Sonny.

GoldenCinderblock: "I keep spending all my money on exotic fish so my armor sucks. Is it possible to romance multiple females? I got with the blue chick so far but I am also interested in the electronic chick and the face mask chick."

    • 3
    • 1
Mar 29, 2017

as always, thanks for your response on the topic @Frank Quattrone" !

Christie Lindor
Management Consultant | Author, The MECE Muse

Mar 29, 2017

robots will at some point replace minimum wage workers in bulk. once this happens people should be a lot more worried about whats next. even robots replacing minimum wage workers could cause some type of revolt which would push back the time-frame for more skilled labor being replaced.

twitter: @StoicTrader1

"He profits most who serves best"

"Every failure brings with it the seed of an equivalent advantage"

*Meditations 8:45

"No one is more hated than he who speaks the truth"

stoictrader.net

    • 1
Mar 29, 2017

The mightiest empires to ever exist, the Romans and the Mongols, relied on this thing called decimation. Perhaps we ought to think about that for once. Dan Brown-Inferno style?

GoldenCinderblock: "I keep spending all my money on exotic fish so my armor sucks. Is it possible to romance multiple females? I got with the blue chick so far but I am also interested in the electronic chick and the face mask chick."

    • 2
Mar 29, 2017

@BEAST MODE TRADER" I think you are spot on with that observation!

Christie Lindor
Management Consultant | Author, The MECE Muse

Mar 29, 2017

I will start to consider being concerned when I have a conversation with an automated operator that does not result in me asking for a human in the first minute.

Only two sources I trust, Glenn Beck and singing woodland creatures.

    • 2
Mar 29, 2017

What happens when AI creates its own version of drugs and liquor? Right now we're in a low-shit system, but once AI gets drunk and on drugs we'll be in a full blown shitcane. You've been warned...

I AM THE LIQUOR

    • 2
    • 1
Mar 29, 2017

That is the catch 23. Long stories get short, for the gooder of us all, we need to stop the worst case Ontario from happening. Otherwise, it will look like a tropical earthquake blew through here.

Only two sources I trust, Glenn Beck and singing woodland creatures.

    • 3
Mar 29, 2017
Jim_Lahey:

What happens when AI creates its own version of drugs and liquor? Right now we're in a low-shit system, but once AI gets drunk and on drugs we'll be in a full blown shitcane. You've been warned...

Im watching you like a shit hawk.

    • 1
Mar 29, 2017

nope

/thread

Mar 29, 2017

Or you could be like Elon Musk with the launch of Neuralink, a brain chip that seeks to infuse technology into people's brains to make them smarter than the robots.

http://www.theverge.com/2017/3/27/15077864/elon-mu...

Christie Lindor
Management Consultant | Author, The MECE Muse

Mar 29, 2017

Yes OP you're absolutely correct, more correct that I ever thought anyone could be. Companies as of last night collectively decided that they will never need financial advice, will never pursue inorganic growth strategies, and will never need to access capital markets.

Here I thought I had a solid job going into banking but guess not. Should've stopped being a prestige whore and taken than janitorial job I guess

    • 1
    • 1
Mar 29, 2017

@therealgekko I never said anything about being right or wrong. I posed the question in the spirit of sharing and exchanging of thoughts and ideas based on 2 interesting books that I have read lately. Thanks for taking the time to respond and adding to the conversation.

Christie Lindor
Management Consultant | Author, The MECE Muse

    • 1
Sep 22, 2017

Photonic transfer will be the advance that moves AI into full gear. Once we are out of binary into stacked information and out of electrons and sending information by photons, the amount of information will be staggering and the speed transformational. Now the data still has to be stored in something, but I see a 50 Year gap in how long it will take before I can buy a personal house bot!

"All men are alike in their dreams, and all men are alike in the promises
they make. The difference is what they do."-- Jean Baptiste Moliere

Oct 31, 2017

[...]

Oct 31, 2017

This is the ideal world idea of automation. We will all have income and free time to achieve the top of maslows pyramid. Art, travel, enlightenment.

This can easily be shown to be drastically optimistic simply by looking at how the vast majority of people spend their free time now - playing angry birds and trolling on yahoo message board.

People are getting dumber at precisely the time where information is freely available to everyone.

Masters in Finance HQ - The #1 site for everything related to the MSF degree!
MSFHQ

    • 1
Oct 31, 2017

I agree Some people will inevitably slack off in education and be unproductive members of society. While many others will still find value in being productive.

    • 1
Oct 31, 2017

Yeah it is an interesting time. It can either be really good or really bad - I don't think there is an in between. If it is good, it will be quite the paradigm shift.

Oct 31, 2017

Good read, what do you think abut automation?

    • 2
Oct 31, 2017

It's a delicate knife. Used well, you can make a fine meal with it, clean it up, wipe it down, and put it back into the block. Used poorly, you can turn the streets red with mass homicide or end up falling on it yourself.

This is a hot topic that Elon Musk bans from discussing during hot tube parties along with talking about if this life is a simulation or not (Neil deGrasse Tyson also thinks this whole game of life is one grand simulation).

For a full background, we have to observe how John Maynard Keynes (1883-1946) thought about the future during his times. This Forbes article summarizes this pretty well.

According to Keynes, we are supposed to be doing the following:

"For many ages to come the old Adam will be so strong in us that everybody
will need to do some work if he is to be contented. We shall do more things for
ourselves than is usual with the rich to-day, only too glad to have small duties
and tasks and routines. But beyond this, we shall endeavour to spread the bread
thin on the butter-to make what work there is still to be done to be as widely
shared as possible. Three-hour shifts or a fifteen-hour week may put off the
problem for a great while. For three hours a day is quite enough to satisfy the
old Adam in most of us!
"

Interested individuals can read his entire essay Economic Possibilities for our Grandchildren listed in the reference below.

TL;DR: We're supposed to be chilling on a 15-hour work week because we still need to be amused and challenged. Why the extra 25 hours (40hr week) or 85 hours (WSO typical) remains is a modern day mystery. The markets are brutally efficient that's why and consumption kept pace with every incremental gain we received by our extra production. If consumption were stifled and our production exceeded this consumption rate, it wouldn't be surprising that we receive days off from the office and factories.

School of Life's How to Make a Country Rich explains our present situation pretty well:

What do I think?: We need automation to keep our prices low so that we can be fat, dumb, and happy glued to our phones or VR headsets. I'm leaning towards a WALL-E end game.

Reference:
Economic Possibilities for our Grandchildren (1930) - John M. Keynes

    • 6
Oct 31, 2017

Well put, one of my favourite videos - I sent it to all my friends. Also you write very logically and clearly. Keep up the good work

    • 1
Oct 31, 2017

Shameless BizInsider/TED Talk/Harvard Business Review-like plug alert

Thanks man! Wrote a thesis in grad school that was well read and I now actively blog.

Glad to keep contributing to the forum and seems this forum can handle my writing just fine.

    • 1
    • 2
Oct 31, 2017

very interesting post

    • 1
Oct 31, 2017

Since you probably researched this topic well enough I'm going to ask.

If you look on US unemployment history chart (1) you can observe that it oscillates around 7.5%, with peaks during panics and recessions. During this time (1890 - 2010) there has been a lot of scientific discoveries, industrial revolution, PC was invented, internet etc. Yet all this failed to create structural unemployment, maybe some short term local lay offs, since replacing a bunch of guy's at a factory with robots is not structural unemployment it's technological modernization. So! Why do you think that automation (and I guess here you refer to AI which is faaaaar from being even close if even possible to conscious human mind) will create massive unemployment? I'm curious. Thank you.

1) https://www.economicshelp.org/blog/wp-content/uplo...

You killed the Greece spread goes up, spread goes down, from Wall Street they all play like a freak, Goldman Sachs 'o beat.

    • 1
Oct 31, 2017

Great question! Here's what I think:

Humans creating machines vs. humans creating machines that create themselves.

A hundred years ago, machines haven't yet started to create or learn for themselves and we were still the architects. Now, that's changing and that's the biggest difference. Machines and machine learning or AI is surprising us with their solutions. First by DeepBlue and Kasparov's defeat and now AlphaGo conquering in a much more sophisticated game. We didn't have powers like this back then and tools that challenged the intellect of its creators.

Speed of information didn't used to double at the current rate back then and information not so readily available. Retraining with less total collective human knowledge was easier relative to the history lesson of today plus getting to the current frontiers of each industry. Check Physics articles and how many authors are required now to publish the next best big discovery vs. Einstein/Feynman/von Neumann era. It's alarming that schools before college haven't really adequately adopted to prepare students to face this "drink from a fire-hose" experience of the real world.

The population game (human capital market) is about to experience a shift as well as baby boomers retire in USA and other countries experience negative population growth.

Lastly, unemployment vs underemployment. HBS grad serving Ben and Jerry's is hardly considered employment to us but to BLS it is perfectly fine! I stopped trusting the BLS number a while ago knowing how job "creation" is defined for statistics reporting purposes. Let's say you are an internal hire, deal is done, you got the job (congratulations). We still have HR float the ad online and then give it the minimum amount of time to stay up. That counts as a job created even though you just changed desks or upgrade your name plate and remain in place.

Oct 31, 2017

Alpha go is simply using probabilistic approach and reinforcement learning to maximize it's evaluation function during the play. This is very far from consciousness. Regarding the "powers we didn't had", we did had these algorithms for more than 50 years. Neural nets were created in 1940, Monte Carlo method in 1930 and Markov chains are more than 100 years old. What we didn't have was computation power to leverage these methods, we have that now (Applause for Nvidia!). Yet again all this is very far from creating a true AI. Current AI applications are good at predefined tasks, yet they cannot think, feel or reason.

Lets conduct a thought experiment, suppose a factory produces bags, they have a certain demand for bags. Suppose they decide to automate production and fire some workers. Well this raises unemployment and decreases consumption, then who will buy the bags if the one who were supposed to buy them do not earn income any more? You can say that automation will decrease production cost thus attracting additional demand. Ok, then other producers follow the first one, automate their production, fire workers, add to supply market shifts to equilibrium we are back where we started only now there is more unemployment. These are only theoretical assumptions but if you look at our history nothing like this happened, because demand has remained in place to consume the products created by automation. Yes robots displace some jobs locally but many new industries emerge that require additional work force.

There might be shortage of necessary skills for a while but that will disappear in one generation time. The worst that can happen is that your grandchildren will need to learn computer science starting from grade one. Actually in my country this will happen from 2018, computer science becomes a compulsory subject in school. So keep calm and learn to code.

You killed the Greece spread goes up, spread goes down, from Wall Street they all play like a freak, Goldman Sachs 'o beat.

    • 1
Oct 31, 2017

Human language, Mandarin Chinese being compulsory soon as well?

The next frontier is the barrier of language and communicating directly via thought and not spoken/written language. We think the same thing (MRI scan) when looking at elephants crossing the road, reading it, or hearing about it. Skip the middle-man, the language processor, and go straight to the source which is thought itself. That would increase our own natural processing speed which has implications both good and bad.

Check out this Documentary called Lo and behold:

Oct 31, 2017

The reason that this is not as big of a problem as people make it out to be is simple: The population is going to decrease to the level supported by automation.

Although the disruption and loss of jobs will probably have painful mid-term effects on our generation, it will only take about 25 years for these to shake out, as old people die off, and younger people choose not to have kids which they can not vigorously support.

Income inequality will likely be exacerbated though, and continue far past this brief adjustment period. However, I do not think inequality is a real issue, rather one manufactured by politicians (like most social problems ver the last 20 years or more). Unless you can prove that the lower end of the spectrum is being actively and legally discriminated against in some way, or at least that their plight is literally and definitely worsening, not relatively worsening when compared to the most successful individuals in the world who make millions of people's lives better in dozens of ways, then I say it's total BS.

    • 1
Oct 31, 2017

Universal income to balance out inequality?

Oct 31, 2017

No thanks, I like my capitalism without a side of communism.

    • 2
Oct 31, 2017

Lol. And this is where my friend, the arm chair student, ran to when pushed for solutions.

Give me, give me mentality to stave off the onslaught of being structurally unemployed forever.

Sorry, Bernie can't help you on this one...

Oct 31, 2017

I really liked your point on the notion of 'Old Adam' when confronted with the idea of automation.

It is my belief that without hard work or achievement, leisure becomes mere boredom and relaxation becomes anxious idleness. It's why I think the idea of "automating away employment" is a quixotic (or perhaps dystopian) idea at best. Humans will always want to be occupied by some pursuit. It's in our nature to create, build and administrate and what separates us from other sentient life forms.

I think several hundred years from now (given humanity is around), people will look back at this era and realize we were living in the beginning of the second renaissance. Old world notions from work to marriage, to politics, to status and wealth are all being challenged and refined into an era of new perception and thinking.

Couple that with the fact the internet has given a voice to the unsophisticated and down-trodden and we see two themes broadly arising: there will be continued friction and unrest in society as the 99% can make their plight known and recognized, and because of that, a much more profound understanding of the human condition on an individual level will lead to a new sense of broad human empathy and societal reform. This is already exhibited by the millennials and Gen Z, the SJW crowd, the rise of the "lifestyle" brand, etc.

Where that will lead when combined with automation (and theoretically more leisure time) is anyone's guess. I just know that despite all the BS, it's a wonderful time to be alive with opportunity abounding for those who chase it and never give up.

    • 1
Oct 31, 2017

This Renaissance - back to basics; a search for meaning in life:

Man's Search for Meaning https://www.amazon.com/dp/080701429X/ref=cm_sw_r_c...
Great sub 200 page read about a Holocaust survivor (Frankl) that can be elegantly summarized by Nietzsche when he said:

"He who has a why to live can bear almost any how."

    • 1
Oct 31, 2017

I like your intellectual ventures into the abstract, Ebolamonkey. We'll get into some heated yet intriguing debates in the future, I presume

    • 1
Oct 31, 2017

Debates create meaning. Differences in opinion makes markets. "I think therefore I am."

Oct 31, 2017

I'm hoping for some ex-Machina snoo snoo before our robot overlords deem us irrelevant.

Be excellent to each other, and party on, dudes.
    • 1
Learn More

2,037 questions across 209 consulting firms. 11 Cases developed by a McKinsey Associate, 10+ hours of video. The WSO Consulting Interview Prep Course has everything you’ll ever need to ace your consulting case interviews. Learn more.

Oct 31, 2017

I'm personally not in ER, though I turned down an internship position in it a few years ago, but when I saw this thread I wanted to share this news article which is recent -- http://www.zerohedge.com/news/2017-03-28/equity-re....
(Yes, I realize its ZH but the actual article is behind a paywall and ZH does a ok job at summarizing it).

    • 1
Oct 31, 2017

Thanks! I read that report and had actually read it earlier in the week. A second time was beneficial! I'm not too sure I like this change. I think I'd really enjoy research -- I find this all fascinating. However, I would not like to join a division that is going to be slowly dissolved, unless of course that I am the best. For example, take Adam Parker, former Chief US Equity Strategist at Morgan Stanley. IB firms still need guys like him, correct?

Oct 31, 2017

I'm sure they do, but like I said I'm not in the ER field nor AM field. My hunch / inference from reading reports and speaking with a Quant PM is that currently and going forwards math & comp-sci may better prepare someone for becoming a PM rather than going from ER --> PM. There have been some interesting academic studies put out by some Quant PM's in finance trade journals you should look into if interested.

Oct 31, 2017

That ZH article is pretty crazy. Thank god I'm not in ER, it's going to be a bloodbath.

    • 1
Oct 31, 2017

I remember reading a book about the the blockchain about a year-and-a-half ago. In that book, the authors described what the taxi industry would look like in the future when there are no drivers and the blockchain is more sophisticated. Basically, each taxi would be its own separate company, with an AI responsible for driving the taxi and taking care of any other business processes. Since AIs theoretically don't care about obtaining profit, taxi fares would be rock bottom. Taxis would also be able to communicate with another, and certain taxis will go into "hibernation mode" if the AI determines that it won't be able to cover its cost based on supply and demand trends.

If such a future comes about, it would make sense to levy a tax on robots.

    • 1
Oct 31, 2017

Mind saving me the Google search and telling me what book that is? That's absolutely fascinating futurology there.

Oct 31, 2017
Oct 31, 2017

This is a tax on capital which will eventually happen. Sadly, it will only give the government more power as people will increasingly depend on Uncle Sam.

Masters in Finance HQ - The #1 site for everything related to the MSF degree!
MSFHQ

Oct 31, 2017

Probably one of fhe few liberal ideas I agree with though. This is not about people being too lazy to work or trying to milk the system. This is about robots doing a lot of jobs cheaper and more efficiently than humans do. It's going to be the industrial revolution times 100. I think displacing a huge chunk of the work force without creating some sort of social safety net is just bad economics. Plus, a lot of these jobs will be jobs that can't be outsourced, so it's not like we're going to lose business because China doesn't have the tax and their cost of production is lower.

What other alternatives are there to handle a situation where such a large chunk of the work force gets displaced? I'd love to hear a free market oriented idea that works.

    • 2
Oct 31, 2017

Agree that is will happen. My concern is with the fact that the government will be the provider and therefore the controller of vast numbers of people. We'd need a bipartisan support for this so it doesn't become one party holding the purse strings.

We need a negative child tax credit. I do not understand why we financially encourage reproduction. The world has far too many people as it is and if anything, on the verge of this automation, we need to discourage population growth.

But I am a malthusian so whatever.

Masters in Finance HQ - The #1 site for everything related to the MSF degree!
MSFHQ

    • 1
    • 3
Oct 31, 2017

Hey man. The reason we financially encourage reproduction is because our entire economy is a giant ponzi scheme. We run deficits year after year and get more progressive in our government spending, social security is one of the programs which WILL be insolvent in decades, or sooner. We need a working tax base to take our older generations debts.

Now this is cynical, but I don't know how else to see it. I doubt this is a master plan or anything by some NWO, but our government is STUPID. I love when people think J.M. Keynes and say he said running deficits promotes economic growth but leave out he was a proponent of austerity and surpluses in good years. O.K. that came out of nowhere but when you're on a roll....

    • 2
Oct 31, 2017

We USED to need a base of workers. With robots you will have to tax capital or production, because you won't be able to cover expenses otherwise.

And once this happens, once robots become articulate enough to do most jobs and people have infinite free time to reach maslow's apex (aka play angry birds all day and eat processed foods) we will no longer have a need to encourage reproduction.

Masters in Finance HQ - The #1 site for everything related to the MSF degree!
MSFHQ

Oct 31, 2017

Geez. The issue is that it's impossible to determine what robot replaced a job, what robot made a job efficient, etc.

Taxing robots would add so much more regulation/lobbying efforts over semantics it would immediately nullify any fairness or efficiency gains due to an alleged distortionary incentive to hire robots over people (there are plenty of incentives to hire people over robots, too).

Right now if you hire a robot if it increases your profits you pay more corporate tax, or whoever sold you the robot pays it. We don't need to make it more complicated.

Oct 31, 2017

Seems to me you just move production off-shore then. Why bother housing your robots in a taxed locale?

Oct 31, 2017

This is interesting. What kind of tax would this actually be? There are many ramblings about moving towards a consumption tax and eliminating the federal income tax. Effectively, under that system, there would be no way to tax robots given that they don't consume anything. Granted if costs go way down in manufacturing because of elimination of humans, then that might help, but it will forever remain inefficient later on when the world catches up.

"Loser terrorists" & "bad hombres"

"Typical candidates are those who attended a top-tier academic institution"
-Most job applications

Oct 31, 2017

You can't tax robots that is an idiotic idea. You can only tax the production.

Follow the shit your fellow monkeys say @shitWSOsays

Life is hard, it's even harder when you're stupid - John Wayne

    • 1
Oct 31, 2017

In a modern context it kinda sounds like an innovation tax.

Oct 31, 2017

Interesting sounding idea, but think it makes little practical sense.

1) Company profits are already taxed and will go up as robots replace more expensive workers. If you want an efficient tax you could just raise corporate taxes and use it for more social welfare
2) B/c a human would need to be paid $50K to do it does not mean the job is 'worth' $50K. You could potentially find another person willing to do it for $25. A tax on lower-cost production doesn't really make sense - once again, just tax profits and ENCOURAGE innovation, which will then drive more profits

    • 1
Oct 31, 2017

It will indirectly target the companies that use these machines for their work purpose.

Oct 31, 2017

Link? Very interested to read the article

Oct 31, 2017

Sorry! Forgot to include it in the original post. Here it is.

Oct 31, 2017

Interesting. For the financial analyst part they mentioned Kensho, does anyone actually know how good Kensho is?

Oct 31, 2017

Computer programmers?!?

Oct 31, 2017

This is the original paper: http://www.oxfordmartin.ox.ac.uk/downloads/academi...

While an interesting topic, I am highly skeptical of numbers like "47%" or how they categorize certain occupations. Statistical and machine learning techniques are not magic black boxes that give answers.

The authors used some forms of economic modeling and statistics to get their projections, but of course the validity of their conclusions depends on the inputs. To make my point: refer to page 30 of the paper:

"To work around some of these drawbacks, we combine and build upon the
two described approaches. First, together with a group of ML researchers, we
subjectively hand-labelled 70 occupations, assigning 1 if automatable, and 0
if not. For our subjective assessments, we draw upon a workshop held at the
Oxford University Engineering Sciences Department, examining the automatability
of a wide range of tasks. Our label assignments were based on eyeballing
the O?NET tasks and job description of each occupation. This information is
particular to each occupation, as opposed to standardised across different jobs.
The hand-labelling of the occupations was made by answering the question
"Can the tasks of this job be sufficiently specified, conditional on the availability
of big data, to be performed by state of the art computer-controlled equipment".
Thus, we only assigned a 1 to fully automatable occupations, where
we considered all tasks to be automatable. To the best of our knowledge, we
considered the possibility of task simplification, possibly allowing some currently
non-automatable tasks to be automated. Labels were assigned only to
the occupations about which we were most confident."

Basically, one of the main inputs is the authors' subjective labels about automation. In some ways, this just seems like a fancy survey.

Anyways, I think automation is more likely in domains where rules are well defined. Like in self-driving cars where the rules are based on physics, diagnoses based on biology, accounting based on accounting rules, etc...

But for finance, since the rules often change (because we deal with people) it's harder to see how certain things could be automated.

Oct 31, 2017

@nauprillion thanks for the mention, and for sharing. In the coming years, big data and faster processing capabilities are going to do to white collar jobs what machines did to blue collar jobs, and I think that the scope of this is far larger than most people realize. Investment banking, for example, will be one of the jobs to be left to the machines at some point. (sorry @cauchymonkey). There might still be a need for people who can explain a model to a client, and there will probably be the need for people to do the relationship management jobs that MDs do now, but at some point we will just be able to tell the machine the parameters, and it will take care of everything on the modeling side.

I think people in general overestimate the complexity of their own jobs, and underestimate both the complexity of other jobs and the ability of technology to replace both. Automation will replace 99% of our current workforce at some point. Forty-seven percent seems ambitious for the next decade or two, though. It is much more likely that automation will change existing jobs to a large degree, but not completely eliminate the need for people in those roles.

I also think this particular study is highly unbelievable. A 96% likelihood that receptionists will lose their jobs to automation? I doubt it - the point of a receptionist is largely to provide a human point of contact for visitors. Looking at the other end of the spectrum, it is highly likely, contrary to this infographic, that doctors and elementary school teachers will be replaced by automation. The "Bottlenecks to Automation" section is also ridiculous. One of my friends is already working on a machine that automatically does your makeup for you.

tl;dr Yes, most of today's jobs will become automated at some point, but don't put any stock into this infographic. While interesting and thought-provoking, it is highly flawed and illogical.

Oct 31, 2017

excel modeling can definitely be automated and artificial intelligence is improving but i am skeptical of the hype.

machine learning techniques like neural networks need lots of data and not every domain has enough good data. moreover, in some areas where relationships and laws change (finance), past data may not be as useful.

for example, when fitting some quantitative model for trading a lot of the time u wont use the last 60 yrs of data since some may have lost relevance.

Oct 31, 2017

The business will always need a human element, it's just that as computers perform more of the actual trading functions, that actual human trading ability (call it skill, luck, gut, talent, experience, etc) is less of a requirement for the job. If you can program and code, know your way around complex mathematical concepts and calculations, there's no reason to worry about it in this lifetime. That having been said the worries you're having are way above your pay grade, so to speak. If you have an offer and an interest, pursue it. It's the only way you'll get answers. A great bit of what you want to know is tied into your personal belief system and nobody's going to be able to give you a 100% answer. Sorry I can't be of more help.

Oct 31, 2017

If you have a job, take it.

Here is my recommendation to you, from someone who works in automated trading:
Embrace the technology- a technophobic or resistance to change is a very sure, very quick way to get laid off. Think of yourself as working for the cash equity business, not as a trader. Electronic trading is booming, high-touch trading is shrinking, though it will never go away entirely.

Make connections, keep an eye on your ass, and try to align your skills with another area you could work in. I have met a few out of work traders, some never worked in finance again, some moved to completely different areas and are still making nice money.

That being said, "traders" will always be needed. Someone has to watch the algos, and while it may seem counter-intuitive, its usually not the guys who build them- they have to go build more algos. The role of a typical trader is much different than it used to be. Instead of making moves, you are usually watching screens and adjusting parameters of some automated system. Jokes have been thrown about on our desk about how traders are just janitors of our systems (another term for this is "risk manager").

Embrace the technology and help advance it and you will do well. Like any area, the guys who stick around are the ones who are looking around the corner and trying to get there first, not the guys trying to maintain the status quo.

Oct 31, 2017

someotherguy - who is paid more, the guys who build the algos or the traders (or janitors/risk managers) that watch them? With this new setup it seems like neither of the 2 can point to a pile of cash and say, I made that, now give me my cut.

Oct 31, 2017

no contest, the guys who build the algos.

Oct 31, 2017

Thanks guys. This definitely makes sense. Let me ask this though: Is it realistic to go from one desk to another after a few years of trading if you want to make a move, or are you kind of stuck where you are since you are so far behind the others in another product? To put it another way, is it possible to move to another product later on, or are you limited in your opportunities?

Oct 31, 2017

I dont think its that realistic honestly to switch products basically once you do get the opportunity the later you make the switch the more difficult it will be to do.

As far as automation I would hardly call the traders overseeing the algos janitors. Sell side guys acting as agents would use the algo to break up large orders and trade most efficiently. A market making or HFT algo would serve as liquidity providers to the buyside. In the end it turns into whos algos are better, it becomes a bit of a chess match and the very principles markets operate under are fundamentally changing. Right now I would give the edge to the HFT algos because they have both the hardware and software edge but I think as time goes along different players will be able to fight back at these.

An anology you could think of it is hacking and security. As hackers find vulnerabilities security looks to plug these vulnerabilities meanwhile hackers are looking for new ways to exploit them. The markets are no different except its all one big chess match.

"Oh the ladies ever tell you that you look like a fucking optical illusion" - Frank Slaughtery 25th Hour.

Oct 31, 2017

I actually have no idea who is paid more. Its tough to tell, the "traders" are all down on the main trading floor, while we are on a side trading floor. They tend to dress better, but image matters more down there. They seem to all have big houses out in the nice neighborhoods in the burbs, but they also tend to be in their late 30's, while we are generally a bit younger. There are also cultural differences at play, down on the main floor, there is a lot more pressure to be flashy and have status symbols, more technically and mathematically minded people are rarely like that at all, aside from showing off their new gadgets.

Keep in mind, I am not working on the prop side of things- our business gets paid off commission fees. Though I have also worked in a prop group doing MM, and the traders seemed to make more, but that was a few years ago.

As far as moving desks, I would say that products that are easy to move between are already on the same "desk." IE moving from futures to options, is easy, but they are usually traded on the same desk. Same w/ most types of Fixed income, and equities as a whole. Moving from equities to Fixed Income and vice-versa... that's a tough transition.

Oct 31, 2017

are we talking quant developers in back office middle office roles vs front facing traders? unless youre a algo developing hedgie like james simons or something similar and youre an algo portfolio manager, and thus trader and quant developer in one, its going to be the trader imho.

Oct 31, 2017

I heard the same thing a nine years ago when i graduated college. Market-making is becoming more exchange-based and less OTC which makes it less profitable for dealers but algorithmic trading really has very little to do with it. On the buyside there are still very few "quant PM" jobs relative to discretionary PMs and that spread widened signifigantly in 2008 when most of the quant funds including supposedly "brilliant" quants like global alpha and AQR got destroyed. The fact that most kids on boards like this seem to think that quant funds are the best place to be makes me more confident that discretionary trading will be around for a long time to come.

Oct 31, 2017

any thoughts?

Don't say this in a banking interview:

Which superhero would you be and why? I want to be like Robin Hood, stealing from the rich and giving to the poor - me.

    • 1
Oct 31, 2017

Yes, eventually.

All it will take is 1 company to do it and then it will cascade throughout the industry.

Oct 31, 2017

Most of the big institutions (JPM Morgan Citi) will automate most jobs in the smaller businesses should be safer

Oct 31, 2017

I do AI and Machine Learning on a daily basis and am obsessed with the space. This question gets asked a lot so briefly:

Sell-side AM/PWN - No chance of survival robots are smarter cheap and scale better. Retail flows will die out leaving only HNW flows which will only survive based on the strength of the client-manager relationships.
Quant HF and Prop - Already the smartest people in room. The use of AI/ML is very interesting because there are deep limitations to what we can do just with technology. These guys use human abilities in conjunction with robot abilities (like centaur in chess) and produce some really cool strategies. No reduction in force anytime soon, but consolidation is possible (already happening in Prop space).
Sell-side S&T - Reductions have already happened. Anything vanilla is robot. Anything more complex will not be. Should be the case for a while to come.
Sell-side IB - Reduction in number of lawyers needed because legal AI is a thing. There's an unrelated reduction in # of bankers due to a general decline in IPOs but otherwise no reduction.

Oct 31, 2017

Thoughts on the buyside? ie value investing

Oct 31, 2017

As a general rule, the less mathematical and longer term a strategy the harder it is to do via computer. Value investing, compared to momentum investing, is not mathematical and will be hard to computerize, so good value investors will always be in demand. Same with PE and VC, both are very hard to computerize.

Oct 31, 2017

What do you think about ER on the buyside (like BlackRock, Fidelity, JPM AM, etc.) ?

Don't say this in a banking interview:

Which superhero would you be and why? I want to be like Robin Hood, stealing from the rich and giving to the poor - me.

    • 1
Oct 31, 2017

There are AI like ROSS that read law and answer legal questions using natural language, meaning you talk to it or type in the question as if you're talking to a real person. I imagine someone will apply this technology soon to build a robot that can answer questions about SEC filings. You could also make it answer questions about the data, e.g. what has been the trend in quarterly COGS, are earnings up or down vs prior quarter.

For the near term, humans will still be writing reports and giving their views, the robots will just speed up the process. It'll be a while before a robot can spit out a cogent investment thesis. Most AI-only strategies are glorified trend-following.

Oct 31, 2017

NOOOOOOOOOOOO!! I need a job damnit...

Oct 31, 2017

lol despite the automation of equity trading, BB desks are still littered with Cash Equity Traders and this process has been going on for awhile.

So with bonds, even though it could happen, the process would be even more lathargic, especially due to the customizable nature of bonds. You cant exactly customize 100 shares of Citi, but you can certainly customize convertible bonds with C as the underlying