Nice read but the use of latent risk to explain his choices (both in tech stack and career) could just as easily (if not more appropriately) be explained through opportunity costs and trade offs.
I don't follow. You can imagine that you have two offers: one job pays $80K/year and one pays $100K/year.
Then by taking the first job, there is an opportunity cost of $20K/year, because you can only have one job at a time, and you had a "better" offer.
But there could be a different latent risk in the 2 jobs. So using that model leads you to evaluate the two opportunities differently. You need another concept to explain the choice; opportunity cost alone doesn't do it.
I feel like im in the authors worst case scenario, as a PHP dev coming up on three years experience, im having a hard time even getting interviews at java/.net shops.
I've lost track of the times ive heard "you have a math degree? Not CS?"
Do you have product like op created in his personal time ? I have seen people without any degree get hired in FANG. If you don't have exp in java .net then you need to show ghis to a company by creating public projects.
Hahahaha. I am sure that acroynm came from an Indian and played a joke on unsuspecting people.
Gaandu is a Hindi abuse whose closet counterpart would be asshole in English.
If you are looking outside pure software shops (FANGs and startups, etc...) getting a .NET or Java certification can be useful. I think they are a bit ridiculous, but they help get past clueless HR types and open up the convo to other parts of your resume.
Yep, plenty of mathematicians get hired by FANG and co. If you can do the whiteboard coding interviews, you're pretty much in, at least for junior positions.
In my experience, places that filter you out based on that, are places I would not want to work in the first place. There are tons of companies that would kill for candidates with math background.
Aside from that, you may well be getting filtered by the HR department’s recruiter who is just going down a checklist. Smaller companies/start ups are more likely to have actual developers looking at candidates and more likely to be more flexible in requirements.
That's weird. The reverse would make more sense. Math is an actual study. CS is definitely not a science--unless opinions and bad statistics count as science these days?
I am in complete agreement, but that’s not the way companies interview. It’s a CS world. Strange because many used to regard a CS degree like a consolation prize if you couldn’t RTFM.
As someone who has around 16 years of dev experience, whose background is originally illustration and graphic design, and then picked up PHP and JavaScript and who has been working at a Java shop for the last 12 years, I can say that much of getting the job is a mix of two things.
One is like the other comment pointed out, having personal endeavors that you can show off is huge for showing that you can build something and that language and degree are less relevant than creatively being able to solve problems.
I had built a commercial CMS tailored for artists all in PHP, built my own theming framework, ORM as well as the payment process and all of the visual design for the system, themes and website. Mind you this was from 2004-2006, and had a lot to learn as I went. I say that not to brag, though I am proud of it, but it's a big part of how I got noticed at the company I'm with. It showed that I could handle multiple areas and solve a domain specific problem (this product formulated when Movable Type was the way to get a blog going, and visual artists, including me, didn't understand all the weird unixy stuff like the command line and the PATH, etc).
So I'd recommend finding an itch you'd like to scratch and see if others have that itch as well and build something. Another upside is you are actually making your life easier in some way.
The second factor that led me to joining this company (I still hate Java and .Net, though I understand their strengths), was because of the company culture itself.
Not only had I technically shown that I had value, but they actually saw my background as a good way of avoiding a monoculture where innovation can't happen. The company was young and it was a risk (we were about 20 people then, now over 800), but it was one that I took mainly because the company's cultural and work values mirrored mine.
If you're getting rejected because of your background and the type of degree, you may just not even wish to work there.
I've had interns and hires come in with degrees ranging from math, electrical engineering, and biology to music and even intercultural outreach, and they became amazing software developers. Heck the original creator of normalize.css and svg4everybody was one of those with a very non traditional background.
So also think of the place you'd like to work for.
I'm not saying those companies are wrong per se, they gotta do what they need to, but personally, I value pragmatism, creativity, transparency and flexibility more than boxes that are just checked on a piece of paper.
If you're a strong logical thinker, like your math degree implies, you'll be better off avoiding shops that filter like that. They don't understand that picking up a syntax is secondary and structured thought is the real strength of a possible hire.
I came from the other way around. I only studied CS in college and my cs101 course is where everything clicked back in 1998. I didn't get my degree and took 10 years of IT to realize I needed to be a developer. I jumped to a company for abysmal pay but they took a chance and I proved I could hack it. The transition from .NET to PHP took a similar leap of faith where I wasn't even sure I wanted to take that divergence. Looking back, I feel I made the right choice but in both instances it required very good friends vouching for my skills as my interview chops were frankly abysmal and may still be.
Now I'm looking to jump to JavaScript or Elixir and analysis paralysis has set in. Unlike the previous transitions, I couldn't have these as my primary focus. I have to resort to personal projects or get very lucky to work with Vue from time to time. Proving myself to myself or other people seems much harder without a very current Github profile and gearing up for interviews requires a form of study that's difficult. I use these techniques consistently enough to be highly proficient but it's not easy to keep that breadth of knowledge on hand to answer most interview questions. I feel like this matters more than a degree at all tbh but there seems to be a bit of selection bias if they're specifically commenting on the type of degree you have. When CS is largely rooted in mathematics and a math degree tends to cover way more advanced topics that I feel are largely applicable.
One of the things I quickly learned as I got close to graduation is that the primary problem with a math degree is that everyone thinks a math degree represents little more than extra time doing the math that they know (or are at least aware they've encountered). There's apparently a blub paradox of education.
There's a somewhat similar dynamic with PHP; when people aren't simply outright going by their own language preferences, they may assume that your caliber as a dev is roughly equivalent to the highest quality PHP codebase they've encountered & groked. And not only has there been an awful lot of low-quality PHP code written, the most commonly deployed PHP code probably doesn't fit the most common web MVCish architecture choices that've coalesced over the years.
You can, of course, use both kinds of judgment in a form of reverse-discrimination.
I think the other solution is to get more concrete about the kinds of problems you've studied/solved. If your Math degree brought you to lots of material about matrix & numerical analysis, hey, you studied machine learning foundations. :) PHP? Maybe you're now partly a domain expert in content management systems.
(But also, probably best to add a language Java/C# folks look at with respect...)
I don't know if I can really agree with the risk classifications the author makes in relation to his career.
I would simply sum what the author is trying to say in one word "Complacency". Having a decent well-paying job can blind you to whether you're on a good career path.
A good career path is one where you are challenging yourself, you learn something new everyday and you position yourself to take on new projects or new leadership roles (if that's your thing).
Besides the problems I have with Taleb I'm happy that the OP had the courage to quit his job and that it worked out. Courage is a good charater trait, and in this case it got rewarded.
That's really well written. I didn't know a lot about Taleb until I saw his attack on (now Dame) Professor Mary Beard on Twitter. I have genuinely never seen anyone so arrogant, he lead an attack on her understanding of the ethnicity of Romans in the UK. Mary Beard is a renowned classics scholar at Oxford, and the attack went along the lines of "You know nothing compared to me because I'm a black belt in probability". He had one study that showed little residual DNA to suggest that there were black Romans in Britain. She had a Hackernewsesq response
1) We know one of the Roman governors was from Africa, but don't know his ethnicity
2) Maybe [something about recessive genes I think] the DNA aged out of the record
3) Maybe most of the Romans went home at the end of the occupation!!!!!!! This one made me think of that Feymnan trait of complete skepticism. Taleb seemed to struggle with the idea that there is not much firm proof of anything that happened in Roman Britain, but not having firm evidence to back up your claim doesn't make you wrong.
It seemed to me, and this is only an impression, that the main problem Taleb had with her is that she was a woman, and a smart one too.
After having seen Talebs Twitter feed, I can't take him serious. He might have (had) some good ideas, but I find it really hard to trust the judgment of a person who berates others on Twitter (e.g. he calls Pinker a charlatan) and think hes the only genius around.
And it also seems that people like Pinker don't take him serious anymore (which seems to aggravate him even more).
He strikes me as the sort of person that loves confrontation. He's a raging narcissist, but he's also brilliant and highly opinionated. I find that I often disagree with him, but I appreciate his perspective. I don't much like when he bullies people in a public forum, since his followers will continue to harrass those individuals.
Not so sure about that. I'm using a heuristic method. Since I'm not a statistics expert and sometimes can't properly assess Talebs statements, I ask myself: "Can I trust the judgement of this person?". Looking at the way he feuds with others (feuds between intellectuals don't all include crude insults), I feel that a person with proper judgement wouldn't behave that way. The same goes for his academic standing, which seems to be poor. These are two red flags.
This said, that doesn't mean he won't offer insight. It just that since I don't have time to read everything, I'll use my heuristic method to look for potential places for insight. In this case it would rather lead me to Pinker et al.
> Then my question to you would be why do you feel that? Is it based on evidence?
Well signaling is a thing thats been developed since humanity exists. For me thats evidence enough. If singaling would be inefficent, it would not still be such an important factor in human interaction.
It's the same with job interviews. No matter how smart you are, you better be on time and show some manners. Because if you don't, you'll be asked: "If you are that smart, why don't you understand that you are not getting a job this way?" and frankly, there's not a good anwser to that.
> If singaling would be inefficent, it would not still be such an important factor in human interaction.
That really depends.
If you avoid a bad hire 10% of the time by rejecting everyone who is late then that can be a good trade off for you even though you have rejected many great candidates. There is a significant cost to a bad hire.
On the other hand you can read a chapter of a book at your local library for free and for a minimal time investment and then decide if you want to keep reading it.
I'd rate Talebs work as significantly more influential and insightful than Pinkers although I agree with Pinker more often.
Beard's point seems to be that ethnic and geographic diversity of some sort did exist and the kind of thing portrayed in the show is not entirely impossible, while Taleb's point seems to be that a family of “black” and “white” people would not be a typical one and that these modern racial categories are different from those of the time. The Atlantic article puts this in context and says why the positions are not inconsistent.
He had an additional objection which was that the portrayal of Mediterranean diversity through the choice of a black (sub-saharan) character was extremely condescending - like "if you're not white European, we'll just pick some random dark skinned ethnicity to represent you, doesn't matter, the important point is non-white."
True. But from the parody you very quickly get the idea what kind of person Taleb is, and the text kind of summarises my feelings towards him, so I find it concise.
But you should of course take it with a grain of salt, since I used this word itself to mock Taleb, implying that his work can be summarised in a short Guardian parody piece.
The idea of latent risk is interesting, but it doesn't seem like the interesting part of a latent risk is it's riskiness. It's true that smoking a cigarette a day accumulates your risk of dying early, but the important part is the accumulation. We can see that if we replace smoking cigarettes with eating donuts, and replace dying early with suffering from obesity. Then we've got exactly the same phenomenon---a bunch of small decisions that are rational taken singly, but disastrous when put together. But there's nothing risky about it---if you eat 5 donuts a day and leave everything else about your life equal, you will become obese with probability 1.
We don't have a really great theory of this kind of boil-the-frog-by-inches phenomenon yet (unless Jon Elster has buried one somewhere in one of his many books, which wouldn't surprise me... and honestly it wouldn't surprise me if Elster anticipated a lot of Taleb somewhere, he's done so so much work on rationality over time), but latent risk probably isn't it.
This just seems like an incorrect analysis, not a limit of rationality. Dosage matters. Dosage specifies a quantity and a time frame. One donut per day is fine. Five donuts is not.
One donut per day isn't fine, not for a person who was already consuming maintenance calories. A single dunkin donuts donut hovers around 300 calories. A pound of fat is ~ 3500 calories, so a donut per day is ~35 pounds a year.
The BMI formula is (lbs/height_inches^2)*703. Allegedly, average male height in the us is just shy of 5'10, female 5'4, so let's just run with 5'7, or 67 inches, as a canonical height. So doing a tiny bit of algebra and taking a derivative, you get ~.16 bmi per pound. A person with 25bmi is classified as normal; 30 as obese. That means that in a little bit over a year you can go from normal to obese on one donut a day.
Why would you assume i meant this was on top of maintenance calories? I find it so bizarre that people look at food this way, that the main challenge is to not over-eat. To somehow eat more without incurring surplus calories. The real challenge is to acquire the necessary number of calories to support your activity. Once you achieve that goal, stop eating, duh.
Narrow answer: the maintenance calories thing was implicit in my original comment (in the "and leave everything else about your life equal," i.e., in addition to the stuff you're already eating, which is presumably hovering around some equilibrium).
Broad answer: the number of people who struggle with their weight is pretty good evidence that the main challenge is, in fact, to not overeat for millions of people.
I think the framing of the question is part of why people struggle. People obsess about gym/running and eating certain types of food. All of that is completely unnecessary. Whatever your level of activity is, eat to that goal. Sometimes i’ll be eating something like a donut and someone will say “do you know how many calories are in that??” as if it’s impossible to simply eat a snack and count it towards your daily energy requirements accordingly. Or as if there are two types of food - ones which keep you alive and ones which taste good but contain these bad things called calories which make you fat and which pose an endless battle against temptation. Another thing which people bring up is slowing metabolism which also seems irrelevant since it’s simply the rate of conversion, not the yield. If people struggle i would frame it in terms of willpower - much harder to resist excessive binge eating after a strenuous workout than after something leisurely like a walk.
/rant
While I agree he made the right decision, and I also agree that Taleb has interesting things to say that bear on career choices, I don't think he needed Taleb's ideas on risk to make this choice. In programming, becoming obsolete is not a risk, it is guaranteed. This is just short term pain vs. long term pain. Which is also a difficult decision, but not one you need to have read "Fooled By Randomness" or "Black Swan" to understand.
Losing your job (eventually) if you become obsolete isn't tail risk, it's the most likely scenario. And losing your job right when everybody else is losing their job is also not tail risk, it's the most likely scenario. So, if you're not learning anything new, you aren't at risk from a black swan, you're doomed.
If anything, it relates more to the idea of "Antifragility"; many small shocks strengthen you. The small shock of learning a new technology (and feeling underproductive while you're on the steep side of the learning curve) is the small shock that makes you stronger. Don't let yourself avoid those small shocks for too long.
After reading Fooled by Randomness, I tried to replicate Nassim Taleb's strategy of buying far OTM options in anticipation of a particular stock being outed as a fraud.
I made a killing when a certain blockchain company traded on NASDAQ was investigated by the SEC and soon faced delisting.
With those winnings, I went all in-purchasing near expiry far far far OTM put options, expecting imminent OTCBB quotes.
It took 2 months for it to happen and my put options expired worthless.
So what I learned is that I suck at trading and I should not gamble.
I think before the market crash in the 80s, he bought a bunch of put options for pennies which ended up being probably several hundred dollars.....literally a 100,000% ROI from a blackswan event
he's still living off that FU money....not sure how rich he is but I estimate 20MM
I am in this situation where I really hope I could apply some risk framework to help me make decision.
I recently signed an offer with a company, US bank with large python community. Now I am counter offered with more pay and promotion and internal transfer to a new team. Conventional wisdom is never accept counter offer. but this one I am giving second thought as it is not for original role but transferred to new team. I will be only one in my location and I will be building MVP for business users. it will be exposed to a lot of tech stack but it won't be scale as large as new company.
However to join a new company, there is always uncertainties like team and project. I had bad experience before. It is difficult to find it out through interview only. It seems to confirm this as I find that hiring process is a bit not taken seriously. It is greenfield project but no coding tests. There is no front end expertise in the team so I was only got asked things like "you worked with angularjs right?". Hiring manager is recruiting three altogether which he thinks will be good team.
risk is about managing uncertainties. I hope decision like this can be priced so it is easier to decide.
54 comments
[ 2.3 ms ] story [ 115 ms ] threadThen by taking the first job, there is an opportunity cost of $20K/year, because you can only have one job at a time, and you had a "better" offer.
But there could be a different latent risk in the 2 jobs. So using that model leads you to evaluate the two opportunities differently. You need another concept to explain the choice; opportunity cost alone doesn't do it.
I've lost track of the times ive heard "you have a math degree? Not CS?"
Like the acronym though.
Aside from that, you may well be getting filtered by the HR department’s recruiter who is just going down a checklist. Smaller companies/start ups are more likely to have actual developers looking at candidates and more likely to be more flexible in requirements.
One is like the other comment pointed out, having personal endeavors that you can show off is huge for showing that you can build something and that language and degree are less relevant than creatively being able to solve problems. I had built a commercial CMS tailored for artists all in PHP, built my own theming framework, ORM as well as the payment process and all of the visual design for the system, themes and website. Mind you this was from 2004-2006, and had a lot to learn as I went. I say that not to brag, though I am proud of it, but it's a big part of how I got noticed at the company I'm with. It showed that I could handle multiple areas and solve a domain specific problem (this product formulated when Movable Type was the way to get a blog going, and visual artists, including me, didn't understand all the weird unixy stuff like the command line and the PATH, etc). So I'd recommend finding an itch you'd like to scratch and see if others have that itch as well and build something. Another upside is you are actually making your life easier in some way.
The second factor that led me to joining this company (I still hate Java and .Net, though I understand their strengths), was because of the company culture itself.
Not only had I technically shown that I had value, but they actually saw my background as a good way of avoiding a monoculture where innovation can't happen. The company was young and it was a risk (we were about 20 people then, now over 800), but it was one that I took mainly because the company's cultural and work values mirrored mine.
If you're getting rejected because of your background and the type of degree, you may just not even wish to work there. I've had interns and hires come in with degrees ranging from math, electrical engineering, and biology to music and even intercultural outreach, and they became amazing software developers. Heck the original creator of normalize.css and svg4everybody was one of those with a very non traditional background.
So also think of the place you'd like to work for.
I'm not saying those companies are wrong per se, they gotta do what they need to, but personally, I value pragmatism, creativity, transparency and flexibility more than boxes that are just checked on a piece of paper.
The industry is so self-defeating.
Have you tried Python?
Now I'm looking to jump to JavaScript or Elixir and analysis paralysis has set in. Unlike the previous transitions, I couldn't have these as my primary focus. I have to resort to personal projects or get very lucky to work with Vue from time to time. Proving myself to myself or other people seems much harder without a very current Github profile and gearing up for interviews requires a form of study that's difficult. I use these techniques consistently enough to be highly proficient but it's not easy to keep that breadth of knowledge on hand to answer most interview questions. I feel like this matters more than a degree at all tbh but there seems to be a bit of selection bias if they're specifically commenting on the type of degree you have. When CS is largely rooted in mathematics and a math degree tends to cover way more advanced topics that I feel are largely applicable.
There's a somewhat similar dynamic with PHP; when people aren't simply outright going by their own language preferences, they may assume that your caliber as a dev is roughly equivalent to the highest quality PHP codebase they've encountered & groked. And not only has there been an awful lot of low-quality PHP code written, the most commonly deployed PHP code probably doesn't fit the most common web MVCish architecture choices that've coalesced over the years.
You can, of course, use both kinds of judgment in a form of reverse-discrimination.
I think the other solution is to get more concrete about the kinds of problems you've studied/solved. If your Math degree brought you to lots of material about matrix & numerical analysis, hey, you studied machine learning foundations. :) PHP? Maybe you're now partly a domain expert in content management systems.
(But also, probably best to add a language Java/C# folks look at with respect...)
I would simply sum what the author is trying to say in one word "Complacency". Having a decent well-paying job can blind you to whether you're on a good career path.
A good career path is one where you are challenging yourself, you learn something new everyday and you position yourself to take on new projects or new leadership roles (if that's your thing).
Besides the problems I have with Taleb I'm happy that the OP had the courage to quit his job and that it worked out. Courage is a good charater trait, and in this case it got rewarded.
1) We know one of the Roman governors was from Africa, but don't know his ethnicity
2) Maybe [something about recessive genes I think] the DNA aged out of the record
3) Maybe most of the Romans went home at the end of the occupation!!!!!!! This one made me think of that Feymnan trait of complete skepticism. Taleb seemed to struggle with the idea that there is not much firm proof of anything that happened in Roman Britain, but not having firm evidence to back up your claim doesn't make you wrong.
It seemed to me, and this is only an impression, that the main problem Taleb had with her is that she was a woman, and a smart one too.
And it also seems that people like Pinker don't take him serious anymore (which seems to aggravate him even more).
PS: http://quixoticfinance.com/empty-statistics/ talks about the Taleb-Pinker feud and rebukes Taleb.
Feuds between intellectuals have been a staple for decades.
You are letting personality problems obscure insights.
This said, that doesn't mean he won't offer insight. It just that since I don't have time to read everything, I'll use my heuristic method to look for potential places for insight. In this case it would rather lead me to Pinker et al.
Then my question to you would be why do you feel that? Is it based on evidence?
Many of the worlds greatest scientists and philosophers have been jerks e.g. Aristotle was a massive sexist.
Or look at the long list of geniuses who were unstable like Galois.
Taleb has always put his money where his mouth is - to me that's a better indicator than poor manners.
Well signaling is a thing thats been developed since humanity exists. For me thats evidence enough. If singaling would be inefficent, it would not still be such an important factor in human interaction.
It's the same with job interviews. No matter how smart you are, you better be on time and show some manners. Because if you don't, you'll be asked: "If you are that smart, why don't you understand that you are not getting a job this way?" and frankly, there's not a good anwser to that.
That really depends.
If you avoid a bad hire 10% of the time by rejecting everyone who is late then that can be a good trade off for you even though you have rejected many great candidates. There is a significant cost to a bad hire.
On the other hand you can read a chapter of a book at your local library for free and for a minimal time investment and then decide if you want to keep reading it.
I'd rate Talebs work as significantly more influential and insightful than Pinkers although I agree with Pinker more often.
And here's an article about the debate/noise: https://www.theatlantic.com/science/archive/2017/08/dna-roma...
Beard's point seems to be that ethnic and geographic diversity of some sort did exist and the kind of thing portrayed in the show is not entirely impossible, while Taleb's point seems to be that a family of “black” and “white” people would not be a typical one and that these modern racial categories are different from those of the time. The Atlantic article puts this in context and says why the positions are not inconsistent.
Talebs point is the public is being mislead by taking an incredibly unlikely scenario and treating it as likely and common.
Mary's rebuttal was that it's not an impossible scenario.
It's funny; but it's an uncharitable summary if you're going to use that as a basis for evaluating Taleb's ideas.
But you should of course take it with a grain of salt, since I used this word itself to mock Taleb, implying that his work can be summarised in a short Guardian parody piece.
We don't have a really great theory of this kind of boil-the-frog-by-inches phenomenon yet (unless Jon Elster has buried one somewhere in one of his many books, which wouldn't surprise me... and honestly it wouldn't surprise me if Elster anticipated a lot of Taleb somewhere, he's done so so much work on rationality over time), but latent risk probably isn't it.
The BMI formula is (lbs/height_inches^2)*703. Allegedly, average male height in the us is just shy of 5'10, female 5'4, so let's just run with 5'7, or 67 inches, as a canonical height. So doing a tiny bit of algebra and taking a derivative, you get ~.16 bmi per pound. A person with 25bmi is classified as normal; 30 as obese. That means that in a little bit over a year you can go from normal to obese on one donut a day.
Broad answer: the number of people who struggle with their weight is pretty good evidence that the main challenge is, in fact, to not overeat for millions of people.
Losing your job (eventually) if you become obsolete isn't tail risk, it's the most likely scenario. And losing your job right when everybody else is losing their job is also not tail risk, it's the most likely scenario. So, if you're not learning anything new, you aren't at risk from a black swan, you're doomed.
If anything, it relates more to the idea of "Antifragility"; many small shocks strengthen you. The small shock of learning a new technology (and feeling underproductive while you're on the steep side of the learning curve) is the small shock that makes you stronger. Don't let yourself avoid those small shocks for too long.
I made a killing when a certain blockchain company traded on NASDAQ was investigated by the SEC and soon faced delisting.
With those winnings, I went all in-purchasing near expiry far far far OTM put options, expecting imminent OTCBB quotes.
It took 2 months for it to happen and my put options expired worthless.
So what I learned is that I suck at trading and I should not gamble.
he's still living off that FU money....not sure how rich he is but I estimate 20MM
I recently signed an offer with a company, US bank with large python community. Now I am counter offered with more pay and promotion and internal transfer to a new team. Conventional wisdom is never accept counter offer. but this one I am giving second thought as it is not for original role but transferred to new team. I will be only one in my location and I will be building MVP for business users. it will be exposed to a lot of tech stack but it won't be scale as large as new company.
However to join a new company, there is always uncertainties like team and project. I had bad experience before. It is difficult to find it out through interview only. It seems to confirm this as I find that hiring process is a bit not taken seriously. It is greenfield project but no coding tests. There is no front end expertise in the team so I was only got asked things like "you worked with angularjs right?". Hiring manager is recruiting three altogether which he thinks will be good team.
risk is about managing uncertainties. I hope decision like this can be priced so it is easier to decide.