https://m.youtube.com/watch?v=2Nx-sH0kn6E
my fiance once asked me when she first met me how to describe myself. I sent her this video.
bang bang skeet skeet!
Umm, ok, if you are happy using existing models and coding something up in python - sure. If you want to build your own models - Ph.D is a must
This is also patently false or the masters programs you’re referring to are garbage.
Well, gl, lol let me know when top hedgefunds / google / fb will hire you in top ds spot
Cognitive dissonance is interesting. I don’t know how you want to delineate this since its subjective, but I know people from my program with masters at Google and a large number with simply undergrads from great schools. The PhD track was opened to me and after spending a decent amount of time talking to people I stayed with MS. Nearly everyone on the PhD track has told me they would not be wasting time on that if they weren’t set on being a professor. Interesting you chose google / fb, L Page only ended with with an MS and Zuck never finished undergrad, neither experience is that rare. How many great practitioners dropped out of school to found major start ups? I’m not recommending that, but acting like you need a PhD is overkill. I’ve worked with a number of PhD students that are definitely not nearly as talented as people on the MS track. It’s also interesting that MIT and HVD only offer the MS in the field and most of those graduates are going to some pretty impressive firms.
The reality is it has become clear to me that even talking to PhD’s, the best money is still earned in investment management or by combining domain expertise with a reasonable level of proficiency. I think the biggest surprise upon getting upclose and personal with CS and DS is for all the hype, how little even top engineers are making versus my expectations. Talking to investment management firms, they are all looking for people with domain expertise in traditional investing that also have a real skill set. Most PhD’s can barely tie their shoes outside of their specialization. Kaggle had an interesting survey showing like 80% of DS implementations fail, largely due to lack of domain expertise and I can tell you first hand that a semester into the masters I was getting inundated with “quantamental” roles.
The idea that you NEED a PhD to get good roles in a less traditional industry where you can build your own demonstrative portfolio of work is a self-limiting mindset.
I think you are confused. Most jobs are applying EXISTING MODELS to make sense of Data. I can pretty much guarantee you, you wont get to design your own models, unless you have PhD. There is still a lot of very high paying jobs for MS in DS, but as you mentioned your comp most likely will be in CS range, topping at 800-1mm for VERY TOP contributers
I’ll just shorten my response to say I’d disagree and I’ve even seen as much with MS candidates that are pursuing research options. I also know the vast majority of PhD’s in the field are not making even $800k. In fact, find me a survey or other research showing a median CS PhD salary above $500k.
You are confused again, i never said Median, i said TOP. You can be quantatative research with MS/DS, sure, and make a ton, but that was not my point
Ok, so you’re saying a top PhD in CS is making $800k to $1M building models from total scratch. This does seem reasonable, although I don’t think that summarily declaring that the best job or pinnacle of the field really makes sense since you have people rolling out DS for apps that wind up making that much and making more than that in investment research quantlite roles. It’s simply one niche of the industry and maybe its desireable to some but summarily declaring that the top job seems pretty subjective.
BS being very defensive again!
The fact is, there are so many MS people that most fail to penetrate the highly elitist circle of PhD researchers. Just look at the credentials of people doing the most sophisticated work at any top institution, and this will become apparent. Whether or not it was worth the cost for them to go down this route - who knows. Would they be there without a PhD? Almost definitely not.
(In fact, I was in a university adjacent location where Google opened a campus not too long ago, and one thing the company did was immediately hire the head of that university’s CS department for a research job. That’s the calibre of some of these jobs.)
As for what constitutes a “top job”, in a field like is being discussed above, it should not be subjective at all - most difficult projects, highest compensation, and most responsibility. Who disagrees with that?
Pretty good year (not in order of importance)…
-
Passed Level 2, first time.
-
Pretty big step up in income. Broke through a mentally significant marker.
-
Our twins were born.
-
I’m a watch collector and purchased a ‘grail’ of mine that I’d wanted for years.
If you throw a ball at 70mph in little league it is like 100mph in major league
I know a youtuber who has no college degree but makes 1.6m off youtube making finance videos. Stephen grAham!!! Ftw
Lolz, not even close.
It’s like saying the top jobs in finance can only be held by PhDs. We both know this isn’t true and there are a myriad of MBA types employing them. Maybe the best jobs in the back office. As I pointed out, Larry Page has an MS and his story is far from unique.
The bolded part is comical, like any field those three things are often mutually exclusive. Which Kip also pointed out. Ergo, the discussion. It’s like some BOM saying the top jobs in finance are building out financial reporting systems. Probably pretty hard, probably dominated by folks with PhDs in sumsuch, not the best job. But I’m sure if they hired a guy somewhere one time, that probably makes your entire point unassailable.
BS, you are missing a point. Finance is completely different than scientific fields. As i mentioned, you can be rank and file data scientiest, or you can be data scientist manager or you can work your way up the chain and become AI Department head, BUT you will never get to work on THE MOST interesting and CHALLENGING problems, as this area is exclusive reserved for PhDs. Also, as a solo contributer your comp is LIMITED vs Ph.D. Now, if you go to finance and become some kind of quant trader, sure, but you could have become quant trader with any MS Science degree, not neceserally data science
I get what you’re saying, but I think its a very narrow view of DS and a very narrow view of what are the most interesting and challenging problems. And this isn’t just me saying this. I was given the option of doing the PhD (which is what I was leaning towards) and opted for the MS only after the majority of the department weighed in contrary to your POV. You also need to realize that you don’t stop learning after school. If you land a good role working on a team on an interesting project it’s very likely you wind up with experience that is at least as valuable to projects like that as an extra two years researching some minutia. This was the point repeatedly made to me by the professors I talked to, that by and large if you land a decent role there’s no reason you shouldn’t be competitive with PhD’s by the time they exit. I sit in on candidate presentations all the time and the work they’re doing is interesting, but anybody viewing this as a great divide, I think is a reach.
Now, that isn’t to say there aren’t some absolute freaks of nature out there that did MIT this, Yale that that are obviously going to be in their own realm, but I don’t think the PhD is the differentiating factor in that equation.
Ultimately it’s probably an agree to disagree situation and we might be just talking past each other but I’ve just been given a different POV from my experience. I’m also not interested in the type of work you’re talking about and doubt I would have been a candidate for that with or without a PhD so I don’t really know what else to add, I’m kind of ambivalent about the discussion at this point.
I think part of the problem is data science is not a clearly defined word, like a buyside job is. I think you are both right. I listen to a lot of data science and computer science podcasts these days and it seems the majority doing novel stuff at the tech firms are PhDs. But there is always a wide array of need in the middle and some are domain experts with less formalized training, but still add value. I’d suspect most businesses do not have to develop their own search algorithm. But it is true the ones that do are often the ones with scale and likely pay the most, since it is such a large economic impact.
I think we all know you have to be a PhD to get hired at Rennaisant, but you can still make a ton of money without the PhD at other firms. I’d guess that’s what the DS industry is like