To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/SamuelBosch . You’ll also get 20% off an annual premium subscription.
You are right about so many things you said in the video. If someone wants to work in academia, a PhD is necessary, in industry... Not so much. I currently work in a hospital and have done so for over a decade, I am extremely burnt out and tired of seeing neuro patients everyday. All of a sudden I fell in love with Cancer epidemiology after a clinic day where I saw three patients in whom I established a diagnosis of prostate cancer, what are the odds! They were all there for spinal problems but left with a prostate cancer diagnosis. I became more curious and started researching and reading about the cancer in my locality, but was baffled at how only very little is known about it's distribution here. So I decided to get a second MSc in PH, specializing in cancer epi. It was the most academically fulfilling thing I have ever done, made me feel alive, and decided it was what I wanted to do for the rest of my life. That's why I am getting a PhD, although my employers don't need it or support me to get one, and I may end up leaving the position, I feel like its my purpose. I feel that's what's missing in many people, the right purpose for doing the PhD
That’s a great and logical reason to do a PhD. You actually have a goal and mission for it, which most PhD candidates don’t. Wishing you best of luck and I hope something great comes out of your PhD 😊
Just another perspective. PhD in the US can be a great way to improve your life, especially if you're from a developing country. I'm a PhD student at a lower-ranked US university. Many people from my country and neighboring countries pursue PhDs because master's degrees are too expensive. PhDs are free here, and the stipend is almost as high as the top salary you could earn. After getting a PhD, you can find a job in software or tech, even at an entry-level. Work for a few years, then do enough to potentially work remotely from your home country. Some of my seniors have done this successfully and potentially will retire before 40. So, a PhD can be a smart strategy for people from developing countries.
That makes a lot of sense. In some way, this is similar to my own situation. Wouldn't have been able to do a master's degreee so easily either. And keep in mind, you can also always "master out" :)
@PKperformanceEU OP says he's from a different country. The money he earns and saves in the US is probably enough for him to live a decent life back in his home country.
I did my PHD in physics at MIT. Never once thought about going into academia as a career. It was all about my love of physics and my love for a real challenge. It was hard,but glad I did it.. never once regretted doing it.
I mean you guys also have like several 5 hours exams during your attempt becoming a Dr.. That's also almost unheard of in any other fields except for maybe some physics & chemistry PhDs around the world.
As a generalist who easily gets bored once 90% of a field is grasped, I probably won't pursue a PhD. Perhaps a start-up is the ultimate fulfillment for all generalists...? Looking forward to what you will achieve with Marveri. Good luck!!
@@filipposchelfi1874 a jack of all trades, master of none. In my field of software development for instance, a generalist would be someone who is decent at frontend, backend, devops and so on. Meanwhile, a specialist would be someone who specializes in one of these tracks and has great amounts of knowledge in it compared to a generalist who may know a little beyond surface level stuff only. Also, no dumb questions, never fear when asking, that's how we all learn.
@@SamuelBoschMIT Good Consulting is invaluable. But I feel like, your impact ultimately is not as high as if you were a high profile project manager in a company, keeping track of every part of a project to make it as good & qualitative as possible. Especially for such high level generalists like you, Samuel. People like you who are able to go so deep into specialized theory, I would not call you a generalist per se. More like a hybrid type: Specialized Generalist - which I think, is one of the most impactful & powerful things a human can be/can possess. Because people can be generalist but only understand superficially the things in specialized area, effectively making them pretty incompetent (sorry for the straightforwardness). I am very confident that you will always be able to detect which project was manages/directed by a specialized generalist compared to just a normal generalist. Almost always the former will have much higher quality, since projects most often are implemented to tackle a specific problem.
Here a Math Phd with three post-docs and changed for HPC one year ago. Of what you say I find something True, something completely False. Restarting as entry level in another job, still another place where to live with little money and the burden to work but also learning at the same time is really hard. And I could have went for this job 12 years ago! I find completely false that the 19 years old me would be faster, more productive and learning more. I can safely say that I would completely destroy the 19 years me in any learning/intellectual competition (I am not talking about something I already know, but about unknown things.). Also, I lived a lot of difficulties and the 19 years me would be very weak in front of troubles for which now I stay calm , breathe and think. The 19 years old me would work much less than me, I get used to work much more than what I was. It could be strange, but also in terms of energy my 19 years could not match me nowdays, the only thing it could do it would jumping about 12cm more and running about 1,5 seconds faster 100m. I am working in HPC, yes there are various people who, instead of something like my abstract Math phd, learnt a lot of IT tools and often I feel embarassed they are so fast and I don't know anything. But other time I look at them as retarded, they don't have any critical thinking and vision of the problems. For example now we are training some Neural Networks and they can not grasp the basic of the variance-bias dilemma in training the models. They forgot any math beyond high school, and I am quite sure that many of them are so used to type just characters on a computer that they could not solve anymore neither a one variable study of function with pencil and paper. Also if you aim to enter working in something like "Nvidia Deep Learning Institute" (where I work there are two guys working there), you need a Phd to work there. The point of a Phd is choosing one that can be with transferred outside Academia with minimal effort, in case you go for a plan B. And there are good topic in many Phds Nowdays for this. The problem is finding mentors that guide you. A point that you did not mention is that (independently from the Institute where you get your Phd) not all Research field are goods, many are fancy dead-paradigmas that go on because there are professors with grants, prizes and power to keep them alive and have new pupils to perpetrate the species (at the end of my first post-doc I realized that I was exactly in this kind research community). A reason because many choose a Phd is because most of the work in companies is awful, in particular IT and consultancy companies(these are pure garbage)
As ML Engineer we know math, Linear Algebra, Calculus 1, and numerical optimization and that is enough, unless we are on the research side, we don't need to specialize in math. You don't need to know quantum psychics to drive a car. Your argument of solving with pen and paper is like saying why use a calculator when they can just add stuff. We don't need pen and paper they are plenty of things that make things easy, so we can focus on other important aspects. Our job is to solve real world problems, and fine tune parameters and alter models if needed. We solve real world solutions, a lot of our models are handicapped by the machines itself, due to the amount of data is getting out of hand. so we don't need further math, cuz the problem lies with the limitations of the hardware, or the nature of the problem is hard. We combine many fields in one. We don't need to be expert at every field. Your job is to develop, our job is to deploy, both have their challenges.
@@OmarElghamry1 Of course. I think that I have been misunderstood. I wanted to say that quitting from Academia for real world is really hard. But sometime you can see who has a stronger theoretical background which can help in some situations.
Samuel, a few of your videos may have helped me cross the mental knot I had in my mind regarding whether or not I should quit my PhD to get into an industry oriented position. Your thought process resonates so much with the values I hold dear. Thank you for making me realize what truly matters.
I am very happy to hear that my video(s) helped you with your decision process. And definitely make sure to also get other peoples' opinions, not to end up overly biased with my own :)
I worked in industry for a little, and now I'm starting my PhD in CS this fall. I loved research and did not like working as a SWE or quant nearly as much as those were the two jobs I worked in industry. It's just personal, decision, I always think people should spend time in industry first then decide; if you go back that means you really want it, and you'll probably be a very good PhD student.
@@SamuelBoschMIT The problem with trying industry and then eventually going back to academia is that in many cases a PhD position only presents to you once, being in the right place in the right moment. Of course there are people with outstanding CVs that will have the door opened any time they knock on it, but for most people its more of a take it now or leave it sort of deal. Another question is that possibly a PhD fresh out of uni is going to be behind his alternative self if he instead been working in industry salary wise. But what is the case going to be 10 or 20 years from there? And related to this two scenarios, who would more valued if in the case of switching companies, the PhD or the one with experience? Cause depending how you look on it, the one that has limited himself to work in a company has better soft skills, but in other sense it's the PhD who has built(paradoxically enough) the deepest and most transferible abilities. I see the point of toughening yourself eventually in real world more applied settings, but sometimes it's the industry job that ends up being the poorer and most repetitive experience Really interesting video and topic in any case, a PhD is certainly not a decision to make mindlessly
I used to work for FB. A PhD starts as IC4, an undergraduate starts as IC3. The amount of time it is expected to take an IC3 undergrad to become IC4 is a year. In 6 years, you can become IC6 or IC7, which is staff level eng. If you are a coder, your compensation at that level would be 800K. You do the math :)
Nice! Question: in your opinion, which are the odds to be employed in FB (or Faang in general) just being an undergraduate vs. holding a PhD? From the pov of an European-based AI engineer like me it seems impossible to even get your CV considered if you don’t have a PhD o like 10 yrs of relevant experience in the field.
I can’t speak for Europe and it definitely has become more competitive in the US in the last couple of years. I got laid off myself, together with a bunch of others in my department, so that should tell you something. But when this current fog clears, it will likely be easy to be hired again. It is a lot about luck and your interview prep. I had a master’s degree in math from a pretty weak university, many years of somewhat irrelevant experience, and a few years of relevant experience. It all depends whether you are good with interviews after you are called. It used to be very easy to be interviewed and they still have a ton of openings I believe. I disliked working for a large company, but the compensation was out of this world given the movement of the stock.
I should also add that basically nobody needs a doctorate to code. There might some positions where it is necessary, but all and all it is a waste of time.
Sure. But the staff maybe just in product team, and the PhD is in FAIR, Facebook AI Research or the equivalent. FAIR probably doing new research, while product team applied what's known
@@NinjaKirikoJediTrue, but you do have an option to move around in FB and if you don’t like what you are doing, you can find something you do like. True that some positions in data science might be closed to you if you don’t have a phd. Core data science, for example, requires a phd. This is much less so for coders though.
The biggest reason for me not to pursue a PhD is that you can conduct nearly the same research in a large company. Especially in the US, I see that research in industry is a big deal. It tends to be more applied and often has a direct connection to our day-to-day life. In contrast, academic research is sometimes better for long-term advancements, but by the time it could become relevant, the industry may have already advanced in another direction, making the research irrelevant, or the industry may achieve the same results through simpler means.
Yeah, that makes a lot of sense. People often talk about "academic freedom" allowing you to do reserach in anything you like. This is often/usually not the case. If you really know what you wanna do, then yes. But usually you end up working on whatever the latest grants were for
In my PhD I studied lasers but primarily about what happens at the center of a focused beam. The laser system itself is already built for you by companies in industry. In industry, to be useful you have to know how to build lasers and control them with electronics, how to build optics and components used in laser systems.
For some certain specialized jobs, getting a PhD is a must. I'm planning to pursue a PhD in Condensed Matter Theory to enable me to work on material science projects that I am passionate about.
I am one. CMT has zero things to say about Materials. The latter is still trial and error in the lab. The former studies ideal systems that have nothing to do with real systems...
I think, in general we have to change our mindset that PhD must be a very difficult journey of a person. What is important is to do the stuffs we love, like researching the topic you are interested in. For teachers, it is great to pursue PhD not only for research but also to be updated of the current trends in their field.
Normally students that can do a PHD usually have great grades already and because of that probably can work at top companies. So its PHD vs working at a top company not some average company. PHD students are usually the best undergraduates.
I think the main reason why one should pursue a PhD degree is the passion for studying and researching a particular topic. Any other reason may not be appropriate, as obtaining a PhD degree is a long and very challenging journey. Job market requires other skills; however, generally speaking, a PhD usually has very high cognitive qualities that can be exploited appropriately and successfully with the right amount of humility and adaptation.
I've got a friend with a high paying industry job. A PhD would bump up his salary but not by much. And doing a PhD would cost him >0.5M in foregone wages over 4 years.
Good point. In my case, I already have over 12 years of experience as a Machine Learning Engineer (sometimes also as a Research Scientist in some roles) and a Master Degree in AI, but I would like to become a full-time AI Research Scientist ( some positions in Industry, as you said, require a PhD ), and having a PhD and also from a well-reputed University like MIT, Stanford, Berkeley, Oxford and so on seems like dream to me. Right now I'm revisiting Calculus, ODE, and others related Math subjects, and doing a lot of research in Meta Learning ( my favorite topic ) to apply for a PhD Position. but being from Brazil, having a lovely son and wife to take care of, it's extreme difficult to compete for a PhD position in US or European University. but thanks a lot for your help and your videos you have been posting here.
There are many applied science and engineering fields where you can find employment in industry, and having a PhD in that field is often the key to getting those positions in the first place.Choosing your school, advisor, research topic, and industrial sponsor are critical in this process.Everyone’s situation will be different.I appreciate your perspective.
People usually have to have some industry exposure first to know of such opportunities, especially the industrial sponsor part. I've went to science conferences for most of my PhD until my last year when I decided to go to an industry conference called Photonics West, hoping to make some connections for job opportunities. By then, my research had already been completed and I was writing my thesis, which is really meant to impress the science community, not industry. When I tell people in industry about my research, they look at me as more of a customer who needs their products than someone who can work for and benefit them internally. Ironically to advance my research I would need better technology that nobody has made yet, or lots of funding so I can have access to lots of different state of the art tech in order to cobble together different systems and keep trying them out to see if they work. Since we do experiments at a lab one month in a year at most, and requires a lot of bureaucracy like proposal submissions and discussions years beforehand, there is not enough opportunity to try things out at high frequency.
Any thoughts on industrial PhD? I saw a relevant number of such positions within IBM and Meta in Switzerland, France, Italy and Ireland and a lot more in Germany with German companies (BMW, SAP, Bosch, Siemens etc). I'm pretty sure such pisitions exist in other European countries as well. The deal is: you are a fixed term employee of the company for 3-4 years and a PhD student at some partner university, they cover your fees (if any) and pay you a usually better salary while you do your research, publish your papers etc.
I’m french, in france i never heard about such thing. It’s not impossible, perhaps it’s in some big companies, but most of the time you make a Phd when you are young. When you work you could have some formation but it’s very hard to make a Phd.
I did not like to work in industry. Prepared docs and moved to the PhD. Industry was really stressful and unhealthy. Could not get not any sleep. Mondays were so dreadful. Crazy, crazy. Now I am pretty much my own boss and I can decide what to do. Although startup path is also not bad. Considering it as a great alternative to PhD.
My BIL went to get his PhD for physics at Caltech for 6 years, and failed. He went home and sulked for 2 years in his parents basement, and only when his parents started complaining he got a job at Safeway (canada) as a stocker and worked there for the last 15 years. He was a researcher and did not want to be a professor, there's not many job avenues to make a living.
I am struggling so much with decisions I want to commit to. I am a second year PhD in theoretical nuclear physics at a reputable university (not top 10, but well regarded in the USA at least). I have been purposely trying to pivot my speciality into something building plenty of practical skills, including machine learning applications. In the past year, I have gradually come to close the academic pathway for myself as I have realized, among other things, the lack of location flexibility that comes with postdocs and professorship. It is just too terrible for me to convince myself otherwise. I would like to end up in quant finance one day, but I am struggling with leaving my PhD or not. I have an MSc in physics with a thesis and an upcoming paper to be published. But down the line, I think I might like to do research in industry, maybe in quant research. All in all, I am in a rough patch as I try to figure out what is best for me. Nice video. This captures my dilemma pretty well haha. I agree with the inertia point, I feel that fear whenever I seriously think about leaving for good.
I think it was a fine decision. Maybe I could be a level 6-7 now (am a senior data scientist aka 4) but I get paid well, am respected by colleagues for the research I do, and I don’t really need to let go of doing research (don’t want to). Those higher levels that you could have achieved right now are essentially are career in politics which is definitely not for everyone. It’s all about science at the end of the day!
I realized that academia was a racket 30 years ago during grad school when I noticed that a 30-page paper suitable for "Physical Review" would get submitted as ten 3-page papers to "Physics Letters".
I completely agree! Fantastic video. I am doing masters in Retrieval Augmented generation in AI and going into Phd soon. I would like just to mention there might be some exceptions. AI jobs actually often do require PhD and it is a place where research and industry merge, mitigating the risks of doing a PhD. If you can enter a AI/ML PhD program in the best schools then it is more often than not worth it.
PhD isn't for average people also it isn't for those who loves money because it requires lots sacrifices and time. You don't need a PhD in order to become a CEO or Executive director. However it is for those who are willing to solve problems and advance our understanding on complex subjects. If you don't like to work under pressure and uncertainty, then don't go for a PhD instead get a job and take care of yourself otherwise you will end up like thise who complain on CZcams about their bad experiences on PhD program.
I worked for the military for 6 years and came to the US for a Masters and understood that getting my green card without a PhD was terribly hard in my field (aerospace engineering), I embarqued in a PhD prorgam in very difficult topic (Computational and experimental fluid dynamics) and hopefully I expect to graduate by spring 2025, I will be 39 years old by that time. By watching this video, I understood that I took the wrong train since the beginning by applying for a Master in aerospace engineering, that was a one way path that made my life terrible hard. Thank you for sharing your experience.
What do you think about short (3 Year) PhDs in Industry like its usually the case here in Germany for example. Dont they just neutralize most of your arguments. Especially if you consider that you need them to climb the coorporate ladder here in germany (most higher ups in big companies do have PhDs). If you then compete against anyone who has a PhD you are usually doomed because they will choose the PhD any day. Feels like what you are saying is mostly for america
If the only reason for doing PhD is the passion and joy for research, please be aware that you can do it on your own without being officially a PhD😂😂 Many leading figures in academia have only MAs and are still leading the field.
I know excellent students who did PhDs that did Quality Job Training after their College Research carreer before landing their job outside academia. People who dropped out of their university carreer paths became teachers or Quality engineers. With PhDs you can contribute to the science community. As an excellent student it may be worthwhile.
My dream: work as a AI researcher at OpenAI, Antrophic, Google, Meta, etc… The problem: I am from Brazil, I study in PUCRS, a university no one knows outside of Brazil. My only shot on landing a job like that, is to go through a PHD a get a publication at a high level AI conference. There aren’t many options for me, except the PHD (I already work as a ML Engineer, but that won’t get me anywhere).
One good thing about science or tech PhD programs is that they are often very low tuition versus a master's program. You are often doing lab work and research that is funded by a grant and that covers all course and research material costs and pays you a small salary. One I am familiar with paid students around $40K/year back in the 2010s.
'Nobody cares or knows about that topic' yeah I bet that most people don't know how a phone or a computer work either, but they still use it daily and it plays a huge role i their lives. Some topic are very important to learn about to contribute to society or be a professional in rare, helpful ressources.
Yeah all the breakthroughs in technology aren't done in academia anymore, it's by the industry, I found academia needlessly sterile and slow, more so focused on the quantity of research papers they churned out rather than quality. Load of bs papers as such. I fared much better in the industry and I dare say I gradually started to understand the architecture and the way things worked naturally after a few months as a SWE. Undergrad took 3 years to explain these concepts, because they're super outdated imo. Currently running my own business w a mentor and quitting my job once it takes off lol, glad I didn't take the chance to do a PhD when offered, if not I wouldn't be where I am today. All the best to the people pursuing PhDs though, I still have deep respect for what you all do.
As much as some points are true some big points are missing … if you are working alone .. this comes down to the lab you are working for .. if you are highly technical and looking for challenges PhD such as PhD in robotics does help you .. A masters degree is simply not enough, because you don’t get enough exposure to what’s need in building robots and especially if you want to be in the edge of technological innovation. And taking the responsibility of your project comes a long way. This builds the quality of manager as well. Technology also keeps changing whether you are in the industry or in the academia so everything changes.
A friend is pursuing a PhD in the US mainly for immigration purposes. Yeah, fair, but it sounds unattractive considering the amount of working hours and minimum wages you have to go through
Great video! I believe that today, the job market is changing so fast that planning for your career 20-30 years ahead (as in academic careers) is just impossible.
Probably worth emphasising that this is US specific - here in the UK a full time PhD usually only takes 3-4 years so you _could_ be entering the workforce at around age 25. (though even then, people I know that have done one here _still_ say "Don't do it !" :) - or at least make very sure you know _why_ you're doing it)
Most countries have an produce more PhDs than they need to fill the open university faculty spots, but to some extend that is necessary. We need a filtering mechanism to allow universities to hire the best brains. The rest can work in industry or work as high school teachers.
I would add that to advance at many biotech companies you probably should have a PhD. This includes moving into management. There is a definitely worker versus researcher division at these firms.
As you work in the industry while waiting to get back in academia, you also need to maintain good relationships with your profs. You need those recommendation letters.
Why can't someone get a PhD in something not very niche? In a topic which people care about like astrophysics instead of quantum effects in high-energy ray propagation through intergalactic magnetic fields.
We’ve made so much progress in science that you cannot be a generalist anymore if you wanna contribute to some research field. If you do a PhD on astrophysics , you’ll still have to work on some tiny niche subfield of astrophysics. That is, unless you somehow discover a new theory of everything, but that’s unlikely
What is your opinion on pursuing a PhD in an ML-related field? For example, apart from my genuine curiosity for computer vision which drives me to want to do a PhD, I feel it’s also quite difficult to become an ML Scientist in industry without one. There is also a very strong connection between ML conferences and industry, and many researchers are even becoming startup founders. Some of the other points you’ve mentioned do apply in this context, but not entirely.
I am thinking of the same but have heard both sides, one saying it’s well worth it and others saying it’s not worth it, i’ve seen a lot of job postings for top tech companies (Amazon, Apple, Google, etc.) preferring a PhD for senior roles, meaning it could come in handy later in a career
Well the mony is really a big focus in this video. Doing a PhD, you earn enough money to make a living. And, arguably, it is worth recommending a PhD to someone that would like to do research, but is not sure about his carreer path. If you like doing something, and get payed enough for that to make a living, that sounds amazing, doesn't it? Sure, there's ways to earn more money on the long run and sure, !many! people starting a PhD do this with the wrong mindset. But I'm arguing that you can very well start a PhD and enjoy the process even if you're not commited to your exact carreer path.
I finished my Masters and worked in industry for awhile and hated it, I remembered enjoying the research I did during my masters and wanted to do more of that so I am currently doing my PhD while you mentioned traditional industries liking you to have a PhD you forgot to mention industrial research R&D departments which are many around the states with many many companies needing people for these positions and they require PhDs
I like my industry salary but I like doing research a lot more and I've been out of school for 4 years and graduated with my Bachelor's 6 years ago. My plan is to complete the MS in Data Science program that I'm starting next month, then a Social Data Science program before I pivot into a PhD in Systems by the time I'm 35, partially since these are the part time options that I could find online but mostly because I've always been interested in analyzing systems from an interdisciplinary perspective. It's probably nuts to do part time online Master's and PhD programs while working in industry, but I have a mortgage and my parents to take care of financially 😅 I also want to give myself the time to make sure that I really do enjoy analyzing systems enough to get a PhD in it (which I have for the past 30 years of my life).
Why would you do anything to impress others? PHD is for you and it is done if you like to do that area. If you do phd about quantum and try to get a job in machine learning you would not like your job if you get it.
I haven’t encountered people overestimating themselves with regard to becoming professors, but rather severely underestimating the toughness of the academic job market. There are dozens of PhDs and postdocs for each professor, it’s literally in the numbers as you said in the video!
PhDs have two career paths, Research and Teaching. Having done a PhD, it’s not fun. It sounds very romantic spending time learning and thinking, but at the expensive of living in poverty for 4-6 years. In all honesty, if you like the biological sciences, go to medical school instead. If you want to work in pharmaceutical drug discovery, research, statistics and in any of the pharmaceutical sciences, you must have a PhD
this is kinda also not true. 1. people become prof from mit universities. The actual reason is, that the better paid prof. typically have a better network. If you do a phd with a very renommated physicist, you will have it easier to become a Prof. A prof my friend does his phd with said, that 50 % of his phds become prof. in the usa (Berlin). 2. it is also not true, that a phd is irrelevant to your work field. In computer science there are many fields where you need a phd to get a researchers position in the big four (paid a lot better). Even in theoretical physics it is beneficial to do a phd in quantum computing or physics informed machine learning, statistics etc.. As many good jobs require a phd.
This video came just in time because I am in the process of deciding between a Phd and a masters. In your opnion, are some PhDs better than others in terms of job market? Perhaps doing one in Machine Learning is better than doing one in quantum effects in high-energy ray propagation through intergalactic magnetic fields? Also, as an entrepreneur, what are your thoughts on starting a business as a backup plan in case there are no jobs after getting a PhD? My true passion is AI research, but I also want to stay employed =(
Another great video, at the right time. A question: Can't you do internships on the side that kinda sort out your finances? I'm sure you could have landed internships from Citadel and all, if you wanted.
If your PhD is very applied and you keep your high school math competition skills sharp, then yes. Similarly, you could intern at Google. However, this only works for PhDs who have extremely strong coding skills
I'm assuming you dropped out of the PhD program because there is no such thing as a former PhD, only former PhD student or candidate. I still think you made the right choice. I have a PhD in Civil Engineering with a concentration in Structural Engineering and I do not recommend it to anyone, unless the individual wants to teach, do research, or like constant mental torture.
I absolutely disagree bro you keep using the word job, fuck a job. Job= dead wage slave job to build someone elses dream. I'd rather do research pursuing my interests than chasing money. Money comes from a love of a pursuit not the other way around
Well, if you don't get hired as a researcher after you finish your PhD, unless you have nice trust fund, you will still be forced to accept a job and just work for the money. So might as well be smart about your decisions, keep this in mind, and decide if a PhD is right for you considering all factors
It's better to hire someone who's really smart and able to learn anything than someone who's dumb and already knows how's to do whatever you want them to do.
fully agree I think I will still try to do a phd but after a career in the industry ^^ (i don't feel wise enough to talk sh$t about machine learning yet x)
I am in a similar state of dilemma right now at the age of 21. I will be going for my master's degree soon (starting in September) which is fully funded for me in a different continent so I don't have any financial burdens as such but after that I am not sure if I want to be a researcher so going for a PhD is probably not the best option for me but because of the nature of my degree which is in material science, I think that is the most common option. I want to know if doing an MBA after my master's degree while getting some work experience should be an option I should consider to increase my chances of getting into the industry or will that be again a waste of 2 years of my life? :/
@Samuel Wenn ich bereits einen Master in CS in DACH gemacht habe, würdest du MBB (Consulting) und dann US MBA nach 2-3 Jahren empfehlen, um nach Amerika zu kommen, oder in Europa bei Big Tech/prop shops als Software Engineer anfangen und L1 Visum wechseln versuchen intern nach US? Siehst du allgemein SWE > Consulting? Vielen Dank für deine Hilfe! :)
Und würdest heutzutage eigentlich überhaupt noch einen bachelor empfehlen wenn man in software engineering oder ai development gehen möchte oder braucht man auch keinen bachelor mehr unbedingt?
Hmmmm da bin ich mir nicht sicher. Die meisten top Softwareentwickler die ich kenne haben einen Bachelor. Das bedeutet aber nicht dass der Bachelor der Grund ihres Erfolges wahr.
@@SamuelBoschMIT Und wie würdest du dich persönlich entscheiden wenn du heutzutage mit dem Abitur fertig wärst? Oder in welche Richtung würdest du überhaupt gehen wollen beruflich?
Hey Samuel, thanks a lot for your video! I strongly believe one needs to know both sides of the coin before picking one. In my case, the first computer science program ever in my country started just back in 2018, so I'm part of that first generation of computer scientists. I've been working as a software engineer since I was in senior year, now it's 3 years. Working there I'd figured out I wanted to go further into one specific problem rather than just find one good enough solution and move on to the next issue. Also when I was an undergrad, I had the opportunity to give classes and I ended up enjoying it quite much. Besides, I spent mostly 1 year and a half doing an undergrad thesis, even though it was just required to spent 6 months. Still I could have done more if more time was allowed (I actually won an award for best undegrad thesis). Those are the strongest reasons for me to do a PhD, I want to then come back to my country and be a teacher/researcher. I also was searching a bit about personality types and I found I was an ENTP so teaching and research were career paths that worked out nice for that personality type.
To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/SamuelBosch . You’ll also get 20% off an annual premium subscription.
You are right about so many things you said in the video. If someone wants to work in academia, a PhD is necessary, in industry... Not so much. I currently work in a hospital and have done so for over a decade, I am extremely burnt out and tired of seeing neuro patients everyday. All of a sudden I fell in love with Cancer epidemiology after a clinic day where I saw three patients in whom I established a diagnosis of prostate cancer, what are the odds! They were all there for spinal problems but left with a prostate cancer diagnosis. I became more curious and started researching and reading about the cancer in my locality, but was baffled at how only very little is known about it's distribution here. So I decided to get a second MSc in PH, specializing in cancer epi. It was the most academically fulfilling thing I have ever done, made me feel alive, and decided it was what I wanted to do for the rest of my life. That's why I am getting a PhD, although my employers don't need it or support me to get one, and I may end up leaving the position, I feel like its my purpose. I feel that's what's missing in many people, the right purpose for doing the PhD
That’s a great and logical reason to do a PhD. You actually have a goal and mission for it, which most PhD candidates don’t. Wishing you best of luck and I hope something great comes out of your PhD 😊
@@SamuelBoschMIT thanks 🙏👍, wishing you all the best!
Most men die WITH prostate cancer, not OF prostate cancer. Your "discoveries" were meaningless.
Doing a PhD means getting paid to think for three years, and that's a luxury
Just another perspective.
PhD in the US can be a great way to improve your life, especially if you're from a developing country. I'm a PhD student at a lower-ranked US university. Many people from my country and neighboring countries pursue PhDs because master's degrees are too expensive. PhDs are free here, and the stipend is almost as high as the top salary you could earn. After getting a PhD, you can find a job in software or tech, even at an entry-level. Work for a few years, then do enough to potentially work remotely from your home country. Some of my seniors have done this successfully and potentially will retire before 40. So, a PhD can be a smart strategy for people from developing countries.
That makes a lot of sense. In some way, this is similar to my own situation. Wouldn't have been able to do a master's degreee so easily either. And keep in mind, you can also always "master out" :)
The wisdom of this perspective can not be overemphasized, especially for those who desperately need to harness this path.
Clever!!👏🏽 👏🏽
Retire?!! You mean living a cheap shit life or else you run out of money?! I will be working until 85yo and feel alive and have good money
If you have the need to retire before the age of 40, then you might want to reevaluate your career.🤦🏻♂️🤦🏻♂️
@PKperformanceEU OP says he's from a different country. The money he earns and saves in the US is probably enough for him to live a decent life back in his home country.
I did my PHD in physics at MIT. Never once thought about going into academia as a career. It was all about my love of physics and my love for a real challenge. It was hard,but glad I did it.. never once regretted doing it.
That's great. If you love physics (as I used to too), then it's definitely not a bad idea. Especially if you wouldn't have liked anything else
My reason is to become a Researcher, not a professor. But later I will go into it. Researcher in industry or non profit or gov lab.
Looks like this guy got his PHD in bodybuilding
💪💪💪
Lol, my thoughts exactly
There’s a reason for it: pure and genuine passion for academia
I'm a pure mathematician so a PhD is the absolute minimum to be someone in the field.
Oof
I mean you guys also have like several 5 hours exams during your attempt becoming a Dr.. That's also almost unheard of in any other fields except for maybe some physics & chemistry PhDs around the world.
As a generalist who easily gets bored once 90% of a field is grasped, I probably won't pursue a PhD. Perhaps a start-up is the ultimate fulfillment for all generalists...? Looking forward to what you will achieve with Marveri. Good luck!!
Yup, startup or consulting or similar business domains can be much bettet for generalists
Dumb question: what's a generalist?
@@filipposchelfi1874 a jack of all trades, master of none. In my field of software development for instance, a generalist would be someone who is decent at frontend, backend, devops and so on. Meanwhile, a specialist would be someone who specializes in one of these tracks and has great amounts of knowledge in it compared to a generalist who may know a little beyond surface level stuff only. Also, no dumb questions, never fear when asking, that's how we all learn.
@@revenger211 thank you so much for the answer!
@@SamuelBoschMIT Good Consulting is invaluable. But I feel like, your impact ultimately is not as high as if you were a high profile project manager in a company, keeping track of every part of a project to make it as good & qualitative as possible. Especially for such high level generalists like you, Samuel. People like you who are able to go so deep into specialized theory, I would not call you a generalist per se. More like a hybrid type: Specialized Generalist - which I think, is one of the most impactful & powerful things a human can be/can possess. Because people can be generalist but only understand superficially the things in specialized area, effectively making them pretty incompetent (sorry for the straightforwardness). I am very confident that you will always be able to detect which project was manages/directed by a specialized generalist compared to just a normal generalist. Almost always the former will have much higher quality, since projects most often are implemented to tackle a specific problem.
Here a Math Phd with three post-docs and changed for HPC one year ago. Of what you say I find something True, something completely False.
Restarting as entry level in another job, still another place where to live with little money and the burden to work but also learning at the same time is really hard.
And I could have went for this job 12 years ago!
I find completely false that the 19 years old me would be faster, more productive and learning more. I can safely say that I would completely destroy the 19 years me in any learning/intellectual competition (I am not talking about something I already know, but about unknown things.). Also, I lived a lot of difficulties and the 19 years me would be very weak in front of troubles for which now I stay calm , breathe and think. The 19 years old me would work much less than me, I get used to work much more than what I was. It could be strange, but also in terms of energy my 19 years could not match me nowdays, the only thing it could do it would jumping about 12cm more and running about 1,5 seconds faster 100m.
I am working in HPC, yes there are various people who, instead of something like my abstract Math phd, learnt a lot of IT tools and often I feel embarassed they are so fast and I don't know anything. But other time I look at them as retarded, they don't have any critical thinking and vision of the problems. For example now we are training some Neural Networks and they can not grasp the basic of the variance-bias dilemma in training the models. They forgot any math beyond high school, and I am quite sure that many of them are so used to type just characters on a computer that they could not solve anymore neither a one variable study of function with pencil and paper.
Also if you aim to enter working in something like "Nvidia Deep Learning Institute" (where I work there are two guys working there), you need a Phd to work there. The point of a Phd is choosing one that can be with transferred outside Academia with minimal effort, in case you go for a plan B. And there are good topic in many Phds Nowdays for this. The problem is finding mentors that guide you. A point that you did not mention is that (independently from the Institute where you get your Phd) not all Research field are goods, many are fancy dead-paradigmas that go on because there are professors with grants, prizes and power to keep them alive and have new pupils to perpetrate the species (at the end of my first post-doc I realized that I was exactly in this kind research community).
A reason because many choose a Phd is because most of the work in companies is awful, in particular IT and consultancy companies(these are pure garbage)
As ML Engineer we know math, Linear Algebra, Calculus 1, and numerical optimization and that is enough, unless we are on the research side, we don't need to specialize in math. You don't need to know quantum psychics to drive a car. Your argument of solving with pen and paper is like saying why use a calculator when they can just add stuff. We don't need pen and paper they are plenty of things that make things easy, so we can focus on other important aspects. Our job is to solve real world problems, and fine tune parameters and alter models if needed. We solve real world solutions, a lot of our models are handicapped by the machines itself, due to the amount of data is getting out of hand. so we don't need further math, cuz the problem lies with the limitations of the hardware, or the nature of the problem is hard. We combine many fields in one. We don't need to be expert at every field. Your job is to develop, our job is to deploy, both have their challenges.
@@OmarElghamry1 Of course. I think that I have been misunderstood. I wanted to say that quitting from Academia for real world is really hard. But sometime you can see who has a stronger theoretical background which can help in some situations.
@@leojack1225 yeah I get you, and I think it is not fair, and sorry if I sounded aggressive.
Samuel, a few of your videos may have helped me cross the mental knot I had in my mind regarding whether or not I should quit my PhD to get into an industry oriented position. Your thought process resonates so much with the values I hold dear. Thank you for making me realize what truly matters.
I am very happy to hear that my video(s) helped you with your decision process. And definitely make sure to also get other peoples' opinions, not to end up overly biased with my own :)
I worked in industry for a little, and now I'm starting my PhD in CS this fall. I loved research and did not like working as a SWE or quant nearly as much as those were the two jobs I worked in industry. It's just personal, decision, I always think people should spend time in industry first then decide; if you go back that means you really want it, and you'll probably be a very good PhD student.
I 100% agree. That’s the right way of doing in :)
@@SamuelBoschMIT The problem with trying industry and then eventually going back to academia is that in many cases a PhD position only presents to you once, being in the right place in the right moment. Of course there are people with outstanding CVs that will have the door opened any time they knock on it, but for most people its more of a take it now or leave it sort of deal.
Another question is that possibly a PhD fresh out of uni is going to be behind his alternative self if he instead been working in industry salary wise. But what is the case going to be 10 or 20 years from there?
And related to this two scenarios, who would more valued if in the case of switching companies, the PhD or the one with experience? Cause depending how you look on it, the one that has limited himself to work in a company has better soft skills, but in other sense it's the PhD who has built(paradoxically enough) the deepest and most transferible abilities.
I see the point of toughening yourself eventually in real world more applied settings, but sometimes it's the industry job that ends up being the poorer and most repetitive experience
Really interesting video and topic in any case, a PhD is certainly not a decision to make mindlessly
1:38 NOBODY
I used to work for FB. A PhD starts as IC4, an undergraduate starts as IC3. The amount of time it is expected to take an IC3 undergrad to become IC4 is a year. In 6 years, you can become IC6 or IC7, which is staff level eng. If you are a coder, your compensation at that level would be 800K. You do the math :)
Nice! Question: in your opinion, which are the odds to be employed in FB (or Faang in general) just being an undergraduate vs. holding a PhD? From the pov of an European-based AI engineer like me it seems impossible to even get your CV considered if you don’t have a PhD o like 10 yrs of relevant experience in the field.
I can’t speak for Europe and it definitely has become more competitive in the US in the last couple of years. I got laid off myself, together with a bunch of others in my department, so that should tell you something. But when this current fog clears, it will likely be easy to be hired again. It is a lot about luck and your interview prep. I had a master’s degree in math from a pretty weak university, many years of somewhat irrelevant experience, and a few years of relevant experience. It all depends whether you are good with interviews after you are called. It used to be very easy to be interviewed and they still have a ton of openings I believe. I disliked working for a large company, but the compensation was out of this world given the movement of the stock.
I should also add that basically nobody needs a doctorate to code. There might some positions where it is necessary, but all and all it is a waste of time.
Sure. But the staff maybe just in product team, and the PhD is in FAIR, Facebook AI Research or the equivalent. FAIR probably doing new research, while product team applied what's known
@@NinjaKirikoJediTrue, but you do have an option to move around in FB and if you don’t like what you are doing, you can find something you do like. True that some positions in data science might be closed to you if you don’t have a phd. Core data science, for example, requires a phd. This is much less so for coders though.
The biggest reason for me not to pursue a PhD is that you can conduct nearly the same research in a large company. Especially in the US, I see that research in industry is a big deal. It tends to be more applied and often has a direct connection to our day-to-day life. In contrast, academic research is sometimes better for long-term advancements, but by the time it could become relevant, the industry may have already advanced in another direction, making the research irrelevant, or the industry may achieve the same results through simpler means.
Yeah, that makes a lot of sense. People often talk about "academic freedom" allowing you to do reserach in anything you like. This is often/usually not the case. If you really know what you wanna do, then yes. But usually you end up working on whatever the latest grants were for
Plenty will require a PhD. Perhaps just because they can.
In my PhD I studied lasers but primarily about what happens at the center of a focused beam. The laser system itself is already built for you by companies in industry. In industry, to be useful you have to know how to build lasers and control them with electronics, how to build optics and components used in laser systems.
For some certain specialized jobs, getting a PhD is a must. I'm planning to pursue a PhD in Condensed Matter Theory to enable me to work on material science projects that I am passionate about.
I am one. CMT has zero things to say about Materials. The latter is still trial and error in the lab. The former studies ideal systems that have nothing to do with real systems...
I think, in general we have to change our mindset that PhD must be a very difficult journey of a person. What is important is to do the stuffs we love, like researching the topic you are interested in. For teachers, it is great to pursue PhD not only for research but also to be updated of the current trends in their field.
Normally students that can do a PHD usually have great grades already and because of that probably can work at top companies. So its PHD vs working at a top company not some average company. PHD students are usually the best undergraduates.
I think the main reason why one should pursue a PhD degree is the passion for studying and researching a particular topic. Any other reason may not be appropriate, as obtaining a PhD degree is a long and very challenging journey. Job market requires other skills; however, generally speaking, a PhD usually has very high cognitive qualities that can be exploited appropriately and successfully with the right amount of humility and adaptation.
I've got a friend with a high paying industry job. A PhD would bump up his salary but not by much. And doing a PhD would cost him >0.5M in foregone wages over 4 years.
Good point. In my case, I already have over 12 years of experience as a Machine Learning Engineer (sometimes also as a Research Scientist in some roles) and a Master Degree in AI, but I would like to become a full-time AI Research Scientist ( some positions in Industry, as you said, require a PhD ), and having a PhD and also from a well-reputed University like MIT, Stanford, Berkeley, Oxford and so on seems like dream to me. Right now I'm revisiting Calculus, ODE, and others related Math subjects, and doing a lot of research in Meta Learning ( my favorite topic ) to apply for a PhD Position. but being from Brazil, having a lovely son and wife to take care of, it's extreme difficult to compete for a PhD position in US or European University. but thanks a lot for your help and your videos you have been posting here.
Wishing you all the best with the applications :)
This was interesting from someone who hasn’t been in academia for a while! Learned a lot :)
😊
There's no such thing as a former PhD. Either you are one, or you are not, unless you've had it revoked due to some scandal.
I think it should have been: "former MIT PhD student".
There are many applied science and engineering fields where you can find employment in industry, and having a PhD in that field is often the key to getting those positions in the first place.Choosing your school, advisor, research topic, and industrial sponsor are critical in this process.Everyone’s situation will be different.I appreciate your perspective.
People usually have to have some industry exposure first to know of such opportunities, especially the industrial sponsor part. I've went to science conferences for most of my PhD until my last year when I decided to go to an industry conference called Photonics West, hoping to make some connections for job opportunities. By then, my research had already been completed and I was writing my thesis, which is really meant to impress the science community, not industry. When I tell people in industry about my research, they look at me as more of a customer who needs their products than someone who can work for and benefit them internally. Ironically to advance my research I would need better technology that nobody has made yet, or lots of funding so I can have access to lots of different state of the art tech in order to cobble together different systems and keep trying them out to see if they work. Since we do experiments at a lab one month in a year at most, and requires a lot of bureaucracy like proposal submissions and discussions years beforehand, there is not enough opportunity to try things out at high frequency.
Any thoughts on industrial PhD? I saw a relevant number of such positions within IBM and Meta in Switzerland, France, Italy and Ireland and a lot more in Germany with German companies (BMW, SAP, Bosch, Siemens etc). I'm pretty sure such pisitions exist in other European countries as well.
The deal is: you are a fixed term employee of the company for 3-4 years and a PhD student at some partner university, they cover your fees (if any) and pay you a usually better salary while you do your research, publish your papers etc.
I’m french, in france i never heard about such thing. It’s not impossible, perhaps it’s in some big companies, but most of the time you make a Phd when you are young. When you work you could have some formation but it’s very hard to make a Phd.
I did not like to work in industry. Prepared docs and moved to the PhD. Industry was really stressful and unhealthy. Could not get not any sleep. Mondays were so dreadful. Crazy, crazy. Now I am pretty much my own boss and I can decide what to do. Although startup path is also not bad. Considering it as a great alternative to PhD.
Samee
Are you in computer science?
My BIL went to get his PhD for physics at Caltech for 6 years, and failed. He went home and sulked for 2 years in his parents basement, and only when his parents started complaining he got a job at Safeway (canada) as a stocker and worked there for the last 15 years. He was a researcher and did not want to be a professor, there's not many job avenues to make a living.
I am struggling so much with decisions I want to commit to. I am a second year PhD in theoretical nuclear physics at a reputable university (not top 10, but well regarded in the USA at least). I have been purposely trying to pivot my speciality into something building plenty of practical skills, including machine learning applications. In the past year, I have gradually come to close the academic pathway for myself as I have realized, among other things, the lack of location flexibility that comes with postdocs and professorship. It is just too terrible for me to convince myself otherwise. I would like to end up in quant finance one day, but I am struggling with leaving my PhD or not. I have an MSc in physics with a thesis and an upcoming paper to be published. But down the line, I think I might like to do research in industry, maybe in quant research. All in all, I am in a rough patch as I try to figure out what is best for me.
Nice video. This captures my dilemma pretty well haha. I agree with the inertia point, I feel that fear whenever I seriously think about leaving for good.
I think it was a fine decision. Maybe I could be a level 6-7 now (am a senior data scientist aka 4) but I get paid well, am respected by colleagues for the research I do, and I don’t really need to let go of doing research (don’t want to). Those higher levels that you could have achieved right now are essentially are career in politics which is definitely not for everyone. It’s all about science at the end of the day!
Nice. Was it a phd in data science? What uni?
I realized that academia was a racket 30 years ago during grad school when I noticed that a 30-page paper suitable for "Physical Review" would get submitted as ten 3-page papers to "Physics Letters".
I completely agree! Fantastic video. I am doing masters in Retrieval Augmented generation in AI and going into Phd soon. I would like just to mention there might be some exceptions. AI jobs actually often do require PhD and it is a place where research and industry merge, mitigating the risks of doing a PhD. If you can enter a AI/ML PhD program in the best schools then it is more often than not worth it.
PHD or no PHD, you're looking pretty jacked Sammy - great vid 👍
In the end of the day, being jacked is all that matters, right? :)
PhD isn't for average people also it isn't for those who loves money because it requires lots sacrifices and time. You don't need a PhD in order to become a CEO or Executive director. However it is for those who are willing to solve problems and advance our understanding on complex subjects. If you don't like to work under pressure and uncertainty, then don't go for a PhD instead get a job and take care of yourself otherwise you will end up like thise who complain on CZcams about their bad experiences on PhD program.
5:08 from my experience every university professor I have worked with cares about their students and they do help them grow.
They usually care. But caring alone doesn’t make you a great manager
I worked for the military for 6 years and came to the US for a Masters and understood that getting my green card without a PhD was terribly hard in my field (aerospace engineering), I embarqued in a PhD prorgam in very difficult topic (Computational and experimental fluid dynamics) and hopefully I expect to graduate by spring 2025, I will be 39 years old by that time. By watching this video, I understood that I took the wrong train since the beginning by applying for a Master in aerospace engineering, that was a one way path that made my life terrible hard. Thank you for sharing your experience.
What do you think about short (3 Year) PhDs in Industry like its usually the case here in Germany for example. Dont they just neutralize most of your arguments. Especially if you consider that you need them to climb the coorporate ladder here in germany (most higher ups in big companies do have PhDs). If you then compete against anyone who has a PhD you are usually doomed because they will choose the PhD any day. Feels like what you are saying is mostly for america
If the only reason for doing PhD is the passion and joy for research, please be aware that you can do it on your own without being officially a PhD😂😂 Many leading figures in academia have only MAs and are still leading the field.
Appreciate the honesty Samuel!
for a few months , I have been thinking about it. It helped a lot, thanks, dear Samuel
You're welcome 😊
I know excellent students who did PhDs that did Quality Job Training after their College Research carreer before landing their job outside academia. People who dropped out of their university carreer paths became teachers or Quality engineers. With PhDs you can contribute to the science community. As an excellent student it may be worthwhile.
My dream: work as a AI researcher at OpenAI, Antrophic, Google, Meta, etc… The problem: I am from Brazil, I study in PUCRS, a university no one knows outside of Brazil. My only shot on landing a job like that, is to go through a PHD a get a publication at a high level AI conference. There aren’t many options for me, except the PHD (I already work as a ML Engineer, but that won’t get me anywhere).
Samuel you missed one thing: research in industry. My dream job is to become a research scientist. This requires a PhD.
One good thing about science or tech PhD programs is that they are often very low tuition versus a master's program. You are often doing lab work and research that is funded by a grant and that covers all course and research material costs and pays you a small salary. One I am familiar with paid students around $40K/year back in the 2010s.
'Nobody cares or knows about that topic' yeah I bet that most people don't know how a phone or a computer work either, but they still use it daily and it plays a huge role i their lives. Some topic are very important to learn about to contribute to society or be a professional in rare, helpful ressources.
Yeah all the breakthroughs in technology aren't done in academia anymore, it's by the industry, I found academia needlessly sterile and slow, more so focused on the quantity of research papers they churned out rather than quality. Load of bs papers as such. I fared much better in the industry and I dare say I gradually started to understand the architecture and the way things worked naturally after a few months as a SWE. Undergrad took 3 years to explain these concepts, because they're super outdated imo. Currently running my own business w a mentor and quitting my job once it takes off lol, glad I didn't take the chance to do a PhD when offered, if not I wouldn't be where I am today. All the best to the people pursuing PhDs though, I still have deep respect for what you all do.
Totally agree with u Samuel, pursuing a Phd just because u want to stay in school is absolute bs
As much as some points are true some big points are missing … if you are working alone .. this comes down to the lab you are working for .. if you are highly technical and looking for challenges PhD such as PhD in robotics does help you .. A masters degree is simply not enough, because you don’t get enough exposure to what’s need in building robots and especially if you want to be in the edge of technological innovation. And taking the responsibility of your project comes a long way. This builds the quality of manager as well. Technology also keeps changing whether you are in the industry or in the academia so everything changes.
do it. it's fun. you learn so much, but more than anything it opens doors. teach anywhere in the world.
A friend is pursuing a PhD in the US mainly for immigration purposes. Yeah, fair, but it sounds unattractive considering the amount of working hours and minimum wages you have to go through
Yeah, I know so many people doing that
Great video! I believe that today, the job market is changing so fast that planning for your career 20-30 years ahead (as in academic careers) is just impossible.
Probably worth emphasising that this is US specific - here in the UK a full time PhD usually only takes 3-4 years so you _could_ be entering the workforce at around age 25.
(though even then, people I know that have done one here _still_ say "Don't do it !" :) - or at least make very sure you know _why_ you're doing it)
Most countries have an produce more PhDs than they need to fill the open university faculty spots, but to some extend that is necessary.
We need a filtering mechanism to allow universities to hire the best brains. The rest can work in industry or work as high school teachers.
I would add that to advance at many biotech companies you probably should have a PhD. This includes moving into management. There is a definitely worker versus researcher division at these firms.
But then who will do research, publish papers and advance human knowledge?
People who love doing this and are great at it. Regardless of whether they get a PhD or not, they are the ones who can do this :)
thanks for talking me out of doing a PhD at MIT and Harvard
There is literally no way for me to be paid to do the kind of work that I want to do outside of becoming an academic researcher.
As you work in the industry while waiting to get back in academia, you also need to maintain good relationships with your profs. You need those recommendation letters.
Why can't someone get a PhD in something not very niche? In a topic which people care about like astrophysics instead of quantum effects in high-energy ray propagation through intergalactic magnetic fields.
We’ve made so much progress in science that you cannot be a generalist anymore if you wanna contribute to some research field. If you do a PhD on astrophysics , you’ll still have to work on some tiny niche subfield of astrophysics. That is, unless you somehow discover a new theory of everything, but that’s unlikely
Currently doing a PhD in econometrics for sustainability, using those skills to plan my investment portfolio... and automate it to help me make money.
Totally Correct
Not everyone can think of a PhD because it requires a great deal of talent and creativity.
What is your opinion on pursuing a PhD in an ML-related field? For example, apart from my genuine curiosity for computer vision which drives me to want to do a PhD, I feel it’s also quite difficult to become an ML Scientist in industry without one. There is also a very strong connection between ML conferences and industry, and many researchers are even becoming startup founders. Some of the other points you’ve mentioned do apply in this context, but not entirely.
I am thinking of the same but have heard both sides, one saying it’s well worth it and others saying it’s not worth it, i’ve seen a lot of job postings for top tech companies (Amazon, Apple, Google, etc.) preferring a PhD for senior roles, meaning it could come in handy later in a career
Well the mony is really a big focus in this video. Doing a PhD, you earn enough money to make a living. And, arguably, it is worth recommending a PhD to someone that would like to do research, but is not sure about his carreer path.
If you like doing something, and get payed enough for that to make a living, that sounds amazing, doesn't it? Sure, there's ways to earn more money on the long run and sure, !many! people starting a PhD do this with the wrong mindset. But I'm arguing that you can very well start a PhD and enjoy the process even if you're not commited to your exact carreer path.
I finished my Masters and worked in industry for awhile and hated it, I remembered enjoying the research I did during my masters and wanted to do more of that so I am currently doing my PhD while you mentioned traditional industries liking you to have a PhD you forgot to mention industrial research R&D departments which are many around the states with many many companies needing people for these positions and they require PhDs
I like my industry salary but I like doing research a lot more and I've been out of school for 4 years and graduated with my Bachelor's 6 years ago.
My plan is to complete the MS in Data Science program that I'm starting next month, then a Social Data Science program before I pivot into a PhD in Systems by the time I'm 35, partially since these are the part time options that I could find online but mostly because I've always been interested in analyzing systems from an interdisciplinary perspective.
It's probably nuts to do part time online Master's and PhD programs while working in industry, but I have a mortgage and my parents to take care of financially 😅 I also want to give myself the time to make sure that I really do enjoy analyzing systems enough to get a PhD in it (which I have for the past 30 years of my life).
He will be putting out a workout video by the end of the year lol
I want to, but there is too much fitness stuff already online 😂
It depends on the field.
Why would you do anything to impress others? PHD is for you and it is done if you like to do that area. If you do phd about quantum and try to get a job in machine learning you would not like your job if you get it.
"Nobody" -> that killed me :)
I haven’t encountered people overestimating themselves with regard to becoming professors, but rather severely underestimating the toughness of the academic job market. There are dozens of PhDs and postdocs for each professor, it’s literally in the numbers as you said in the video!
PhDs have two career paths, Research and Teaching. Having done a PhD, it’s not fun. It sounds very romantic spending time learning and thinking, but at the expensive of living in poverty for 4-6 years. In all honesty, if you like the biological sciences, go to medical school instead. If you want to work in pharmaceutical drug discovery, research, statistics and in any of the pharmaceutical sciences, you must have a PhD
If doctoral students looked like you I'd probably be in a program right now 🤤
Great video. Basically, if you want to do research in Academia or Industry, get a PhD. Otherwise, a masters is more than enough.
I am so sorry for you. I hope you find your way!
this is kinda also not true.
1. people become prof from mit universities. The actual reason is, that the better paid prof. typically have a better network. If you do a phd with a very renommated physicist, you will have it easier to become a Prof. A prof my friend does his phd with said, that 50 % of his phds become prof. in the usa (Berlin).
2. it is also not true, that a phd is irrelevant to your work field. In computer science there are many fields where you need a phd to get a researchers position in the big four (paid a lot better). Even in theoretical physics it is beneficial to do a phd in quantum computing or physics informed machine learning, statistics etc..
As many good jobs require a phd.
At MIT and similar universities, you learn how to get grant money, network and get papers published by those who are most successful in these areas.
never overinvest... the pride of your mom is not worth it...
Ok bro, say whatever you want, I'll do it anyway bcs I love science.
This video came just in time because I am in the process of deciding between a Phd and a masters. In your opnion, are some PhDs better than others in terms of job market? Perhaps doing one in Machine Learning is better than doing one in quantum effects in high-energy ray propagation through intergalactic magnetic fields?
Also, as an entrepreneur, what are your thoughts on starting a business as a backup plan in case there are no jobs after getting a PhD? My true passion is AI research, but I also want to stay employed =(
Do you get food with PhD you know how many people get blindness because of watching screens. Those wide high definition TV's
Another great video, at the right time.
A question: Can't you do internships on the side that kinda sort out your finances? I'm sure you could have landed internships from Citadel and all, if you wanted.
If your PhD is very applied and you keep your high school math competition skills sharp, then yes. Similarly, you could intern at Google. However, this only works for PhDs who have extremely strong coding skills
Most stacked doctor
I'm assuming you dropped out of the PhD program because there is no such thing as a former PhD, only former PhD student or candidate.
I still think you made the right choice. I have a PhD in Civil Engineering with a concentration in Structural Engineering and I do not recommend it to anyone, unless the individual wants to teach, do research, or like constant mental torture.
😂
Constant mental torture
I absolutely disagree bro you keep using the word job, fuck a job. Job= dead wage slave job to build someone elses dream. I'd rather do research pursuing my interests than chasing money. Money comes from a love of a pursuit not the other way around
Well, if you don't get hired as a researcher after you finish your PhD, unless you have nice trust fund, you will still be forced to accept a job and just work for the money. So might as well be smart about your decisions, keep this in mind, and decide if a PhD is right for you considering all factors
And who is going to fund pursuing your interests?
Like it or not, money is not very far from your choices
It's better to hire someone who's really smart and able to learn anything than someone who's dumb and already knows how's to do whatever you want them to do.
The free master's degree is nice but what do you need a master's degree for?
But how to get a chance to work in america if we don't do PhD ? Can you give me some advice, please. I am a 4th year cs student
You could just get a master’s degree in a STEM field
@@SamuelBoschMIT yes. But not much master scholarship in computer science in us :(
fully agree
I think I will still try to do a phd but after a career in the industry ^^
(i don't feel wise enough to talk sh$t about machine learning yet x)
If you want to work in academia a PhD is a must, but if you're leaning toward industry, your time is better spent elsewhere
I am in a similar state of dilemma right now at the age of 21. I will be going for my master's degree soon (starting in September) which is fully funded for me in a different continent so I don't have any financial burdens as such but after that I am not sure if I want to be a researcher so going for a PhD is probably not the best option for me but because of the nature of my degree which is in material science, I think that is the most common option. I want to know if doing an MBA after my master's degree while getting some work experience should be an option I should consider to increase my chances of getting into the industry or will that be again a waste of 2 years of my life? :/
Hi Samuel, is doing a PhD in AI worth it if I want to move to research domain?
@Samuel Wenn ich bereits einen Master in CS in DACH gemacht habe, würdest du MBB (Consulting) und dann US MBA nach 2-3 Jahren empfehlen, um nach Amerika zu kommen, oder in Europa bei Big Tech/prop shops als Software Engineer anfangen und L1 Visum wechseln versuchen intern nach US? Siehst du allgemein SWE > Consulting? Vielen Dank für deine Hilfe! :)
Und würdest heutzutage eigentlich überhaupt noch einen bachelor empfehlen wenn man in software engineering oder ai development gehen möchte oder braucht man auch keinen bachelor mehr unbedingt?
Hmmmm da bin ich mir nicht sicher. Die meisten top Softwareentwickler die ich kenne haben einen Bachelor. Das bedeutet aber nicht dass der Bachelor der Grund ihres Erfolges wahr.
@@SamuelBoschMIT Und wie würdest du dich persönlich entscheiden wenn du heutzutage mit dem Abitur fertig wärst? Oder in welche Richtung würdest du überhaupt gehen wollen beruflich?
Excellent video (by a phd drop) !!!
Bro now I am happy to get rejected by MIT
A former Ph.D.? Really? I thought you dropped out?
😂😅😂😅😂
Don't do a PhD because this influencer prefer to hire people with lower education and feel insecure hiring people with PhD.
Bro recommend doing an MBA tho (rofl).
Are you Lemillion from my hero academia?
Hey Samuel, thanks a lot for your video! I strongly believe one needs to know both sides of the coin before picking one. In my case, the first computer science program ever in my country started just back in 2018, so I'm part of that first generation of computer scientists.
I've been working as a software engineer since I was in senior year, now it's 3 years. Working there I'd figured out I wanted to go further into one specific problem rather than just find one good enough solution and move on to the next issue. Also when I was an undergrad, I had the opportunity to give classes and I ended up enjoying it quite much. Besides, I spent mostly 1 year and a half doing an undergrad thesis, even though it was just required to spent 6 months. Still I could have done more if more time was allowed (I actually won an award for best undegrad thesis).
Those are the strongest reasons for me to do a PhD, I want to then come back to my country and be a teacher/researcher. I also was searching a bit about personality types and I found I was an ENTP so teaching and research were career paths that worked out nice for that personality type.
For wvery dont do a B there is a do a B video iut there, for everything. It is all about listeing to different opinions.
You have large triceps and biceps. Must be from having to carry all those heavy books one needs to read to acquire a PhD.