Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Syllabus for Eric's PhD Students

docs.google.com

Post reply on HN
Print view View original

81–90 of 95 posts

Re: Syllabus for Eric's PhD Students

#81

Post by johntfella »

jasonzemos wrote: post #79
johntfella wrote: post #75

Earlier quoted context omitted.

I preferred wandering the woods. I transferred to a university that was more about wandering than coaching. I'm skeptical of the calls to overhaul. Undergrads and even masters degrees went through this and now we have group projects and less individual learning. Too much focus on team work and collaboration. If anything we need to be teaching more wandering through the woods, more being on the out over the in of the…

Or bring back the Habilitation above PhD for serious academics. PhD is already sought after too much for professional reasons. Another 10 year commitment distills the lovers of wisdom from the lovers of cash.

I like this idea. Longer time commitments would filter this out.

Re: Syllabus for Eric's PhD Students

#82

Post by HenryBemis »

ta1234567890 wrote: post #64
HenryBemis wrote: post #60

Earlier quoted context omitted.

Project Management. I once was working with two CS professors in a UK university. One was full of words (let's call him A). The other was one of the brilllian minds in UI science that helped design the Nokia phones' keypads (let's call her B). Most people disliked B because she had a start, middle, end. She required timelines, milestones, deliverables. Because of that everyone working with her was getting better grad…

Project management is definitely very important. As a group/team manager where do you put the "deep thinkers, not doers" though? What is their place in business or society? I personally identify with being a thinker instead of a doer and for most of my professional life struggled trying to find a position from where to contribute without feeling inadequate for not doing/executing.

I don't feel you "push them" to a specific slot. Thinking is good but something needs to come out. A book, a paper, a.. something. A number of people thinking for a common subject need to produce something, not think for 3 months and then start thinking of something else.

The PM skills is to get them to produce efficiently and effectively. Not perpetuate the "sitting and thinking".

Re: Syllabus for Eric's PhD Students

#83

Post by fuzzfactor »

eegilbert wrote: post #5
legerdemain wrote: post #4

Earlier quoted context omitted.

> Test whether the PhD advisor you select knows this and has a history of taking this approach with former students. How do you "test" for this? Asking potential advisors "do you know this?" is blunt and rude. Asking potential advisors in a mannered way prompts anodyne answers. Current students in the potential advisor's lab are unlikely to offer an intimate assessment of their personal satisfaction and research prog…

This is my syllabus. :-) I tell all my prospective students to talk to my former students.

This is brilliant in a way that relates to commercial research in natural science too, generalizing some of the terms and focusing more toward business than academic life:

1. What difference will this work make if you succeed?

a. Whose lives will be made better by this research? -the client

b. How will this improve upon what’s currently being done? -the client will be more satisfied

c. Why is this one of the most important questions in the field? -because a valuable client asked

that was easy

d. Will it create a big policy change at some level (company, government)? -if so it will be for the positive with consensus, otherwise no costly changes

e. Will it inspire a new class of systems? -would be good if it was worth building beyond clients' immediate needs

2. Who will care about it when you’re done? -two clients would be better than one

a. Will government agencies care? -only in a beneficial way, when the FBI comes to your office you want to be their consultant not their suspect

b. Will platforms and industry care? -if you do projects which make them money, more fondness can be expected than if you cost them money

c. Will non-commercial researchers care enough to recognize or benefit from it?

d. Will non-commercial researchers care enough to teach it to their staff/students?

e. Will anyone care about it 10, 20, 50 years in the future?

3. How will this change what other people (defined broadly) are doing?

a. Will other researchers change what they’re working on after seeing your work?

b. Will practitioners do something different?

c. Will users adopt what you’ve made, found, created?

d. Will authorities use what you’ve found to draft new requrements/specifications? -unless no negative impact can be imagined, such a sensitive project might best be shelved until a time of more positive outlook

Professor Gilbert has an _internal deadline_ 7 days more strict than the real publication cutoff, actually allowing for reduced stress and making higher Quality output possible in ways others may never achieve:

>If the paper isn’t ready, we won’t submit it to that conference or journal deadline; we’ll submit it somewhere later.

Naturally deadlines for being AT the conferences are not missed. I bet they can build anticipation about future presentations most accurately when this happens.

Re: Syllabus for Eric's PhD Students

#84

Post by Ar-Curunir »

dude_bro wrote: post #31

I think a point of tension for grad students is that many don't realize a PhD program is more so a job than it is another part of your education. Your advisor is generally paying for you with their hard earned grant money and relies on your results for their own success. In typical 9-5 job terms, the advisor is the manager and the PhD students are the individual contributors. It's a harsh realization that many grad s…

> I think a point of tension for grad students is that many don't realize a PhD program is more so a job than it is another part of your education.

I don’t think grad students fall into that trap; they know that PhD is all about research, and very little about classes.

It’s more the non-academic world which has that misconception

Re: Syllabus for Eric's PhD Students

#85

Post by fuzzfactor »

bonoboTP wrote: post #19
supernova87a wrote: post #2

I will just offer some unsolicited advice about PhDs, as it came up a few weeks ago also. And the author touches on this. Of course this mostly applies to empirical sciences, but maybe some theoretical ones too. An effective PhD advisor/thesis isn't wandering the woods to find something. It's a guided coaching exercise, with an outcome in mind. Test whether the PhD advisor you select knows this and has a history of t…

> Concretely, even if you don't know what the answer is going to be at the end of your research, you must think about, or have an idea about, the format of what that amazing answer is going to be. Write the outline of your thesis and "ghost out" what the major charts will be. Write the intro sentences of each chapter -- what are they? (and I don't just mean the boring review of the field part, but your findings part)…

>In the rarest cases do people come out of it as a positive passionate person. Even the successful ones become a hedging, scheming, lukewarm, slimy academic personality. Everything encourages this. The way you get professor positions, the way you get grant money etc. It optimizes and chisels people into becoming the antithesis of curious, open scientific inquiry.

Got a point there, all kinds of academic pressure other than the push for scientific progress.

There should be more than one way to outperform what you could do in such an environment, or in the industrial counterparts following that pattern.

>Several have given up and went to industry.

I think worked for these bozos before, not fun but milestones are milestones.

>Ideally in an alternate universe of different incentive structures that support real science, a person would work in research, use existing approaches in different scenarios, push and prod them in different use cases then perhaps once every year you'd stumble upon some interesting effect you don't understand. Then you'd dive in and prod this specific effect with well designed experiments. Once it's clearer what is going on, you come up with the best way to convey this newly earned knowledge. It is at this point (once you actually know what you want to alert the research community to) that you decide how to best present the knowledge, what tables, sections and paragraphs will succinctly and ergonomically convey the information.

That's what I did on my home planet.

Had to build my own laboratory.

Works better than you would think.

Re: Syllabus for Eric's PhD Students

#86

Post by yiyus »

pmohun wrote: post #45

This is a fascinating look inside the academic system. Eric, a question if I may: Is there room for a modern PhD that values breadth over depth? The goal of completing a Ph.D. program is to become the most knowledgable person in the world about a specific topic. While this is a valuable goal for disciplines that reward specialization, it does not provide enough flexibility for autodidacts with a range of interests. A…

> The goal of completing a Ph.D. program is to become the most knowledgable person in the world about a specific topic. While this is a valuable goal for disciplines that reward specialization, it does not provide enough flexibility for autodidacts with a range of interests.

This is generally true, but it is not a rule written in stone. One word: interdisciplinary.

For example, I work in a field which lays in the middle between materials science and mechanical engineering. My day to day work is mostly programming, but I also need a strong background on experimental methods and fundamental physics (also, math and statistics, lots of that stuff people say they study but never need in their careers).

We need many people like this. We need mathematicians who can interpret experimental results. We need engineers who can question "fundamental" laws. We need scientists who are good programmers and are not afraid of complicated equations. In my department, jacks of all trades are a very valuable asset.

I am also of the opinion that people with a broad range of interests usually have more original ideas. People very focused in some field do very important work, and find very good jobs when they finish, but important advancement usually comes from thinking-out-of-the-box people.

Re: Syllabus for Eric's PhD Students

#87

Post by chrisseaton »

Beldin wrote: post #67
g_p wrote: post #54

Earlier quoted context omitted.

I felt your way actually about it at the time. I like the 3 year timescale and greater level of self led work. I pretty much avoided doing any "taught" courses for the reasons you outlined, but I have found some of my students did benefit from them. I think it comes down to individuals and what they need. It wasn't right for you or me, but I think there are some students who perhaps come in needing a little bit of a…

I'm not from the UK. The idea of someone becoming an actual PhD in 3 years is rather quaint to me. Perhaps it's just a different approach, or perhaps the PhD system I'm used to produces a level of results that UK PhDs typically only achieve after their first PostDoc. Or, perhaps there is an initial period of a year or so where the student is not yet doing a PhD, but is trying to produce results nevertheless. Saw that…

In the UK you just get going on your research. You don’t do initial courses. You effectively defend your proposal when you apply.

Ultimately the goal of a PhD is to learn to be a researcher and to produce a good new research result. If you can do that in three years why wait around another two or more for the sake of it?

I had a colleague who did a PhD in two years in Austria. In that time he got two top-tier papers published. If you’re getting multiple papers into top-tier venues then surely you’ve past the test? You obviously can do research and you obviously are producing good results as judged by a wide group of peers.

Why does the US drag it out so much?

I’ll tell you why - US PhD students also only spend about three years on their PhD. They spend the rest of their time doing masters-level taught classes, teaching (!) and working on their advisors’ projects instead of their own!

Re: Syllabus for Eric's PhD Students

#88

Post by rleigh »

johntfella wrote: post #75
n3ur0n wrote: post #11

Earlier quoted context omitted.

As a rising 5th year PhD in ML -- I could not agree with this advice more! I have very hands-off advisors. I spent the first 3 years "wandering the woods to find something". Last year, I really had to sit down and think about how I can finish up my PhD on time. I pretty much did what you outline here. A lot of tools I used were organization tools I learned from my business school/product manager friends. I honestly t…

I preferred wandering the woods. I transferred to a university that was more about wandering than coaching. I'm skeptical of the calls to overhaul. Undergrads and even masters degrees went through this and now we have group projects and less individual learning. Too much focus on team work and collaboration. If anything we need to be teaching more wandering through the woods, more being on the out over the in of the…

Many years ago, I would have disagreed with you about the group projects. Today, I'm more in favour of them. The reason for my change of heart is that the moment you step outside acadaemia, be it industry in the same field, a research organisation, or any company, you'll be working on "group projects" as the default way of working. They are group efforts, and learning to work well in a group context is a valuable skill.

One big problem acadaemia has is the individual nature of research, even within research groups. A vast amount of time and effort is wasted on trivia due to working alone, which would be quickly resolved in a group. And while there is value in individual intellectual effort and reward, it can be vastly greater returns in a group context.

None of this is to say that lone investigation of out there ideas is bad. I think that's essential. But not for 100% of your time.

In many research settings, the scale of projects requires being a small part of a larger group. And while groups can squash good ideas, they can also kill bad ideas, and there needs to be a balance.

Not all PhD students are Einsteins, and it's as important to prevent people diving down a rabbit hole which leads nowhere as it is to give them the freedom to explore or else they might find they wasted several years rediscovering something already known or discovering nothing. I saw this happen to one of my fellows as well as myself. A good supervisor should be guiding appropriately. I found to my horror a PubMed search revealed a paper dating back to 1992 which had done several months of my research nearly two decades prior. When there are so many millions of papers, it's all to easy to miss stuff in the noise.

Re: Syllabus for Eric's PhD Students

#89

Post by rleigh »

n3ur0n wrote: post #11
supernova87a wrote: post #2

I will just offer some unsolicited advice about PhDs, as it came up a few weeks ago also. And the author touches on this. Of course this mostly applies to empirical sciences, but maybe some theoretical ones too. An effective PhD advisor/thesis isn't wandering the woods to find something. It's a guided coaching exercise, with an outcome in mind. Test whether the PhD advisor you select knows this and has a history of t…

As a rising 5th year PhD in ML -- I could not agree with this advice more! I have very hands-off advisors. I spent the first 3 years "wandering the woods to find something". Last year, I really had to sit down and think about how I can finish up my PhD on time. I pretty much did what you outline here. A lot of tools I used were organization tools I learned from my business school/product manager friends. I honestly t…

Part of my PhD programme was a "transferrable skills" set of courses which included planning and project management, Gantt charts and all that stuff to enable you to plan several years of work.

Re: Syllabus for Eric's PhD Students

#90

Post by iib »

n3ur0n wrote: post #11
supernova87a wrote: post #2

I will just offer some unsolicited advice about PhDs, as it came up a few weeks ago also. And the author touches on this. Of course this mostly applies to empirical sciences, but maybe some theoretical ones too. An effective PhD advisor/thesis isn't wandering the woods to find something. It's a guided coaching exercise, with an outcome in mind. Test whether the PhD advisor you select knows this and has a history of t…

As a rising 5th year PhD in ML -- I could not agree with this advice more! I have very hands-off advisors. I spent the first 3 years "wandering the woods to find something". Last year, I really had to sit down and think about how I can finish up my PhD on time. I pretty much did what you outline here. A lot of tools I used were organization tools I learned from my business school/product manager friends. I honestly t…

Would you want to list those organization tools--if they are publicly available?

I find that a lot of important nitty-gritty knowledge is dismissed as besides the point, something akin to why MIT designed the "Missing Semester" course [1]. I would have a very hard time progressing with research if I didn't find Zotero [2], for example.

[1] https://missing.csail.mit.edu/ [2] https://www.zotero.org/

Post reply on HN