Hacker News Reader

I think the biggest testament to this is the 50%+ dropout rates common in PhD programs. You're starting with some of the most intelligent, diligent, bright-eyed young people on the planet and letting them work independently in a field that they presumably have a deep passion for.

There's some truth to this, but the other reality is that academia is a pyramid scheme; the scheme worked from 1945 - 75, when higher ed grew rapidly, and has worked more or less poorly for PhDs ever since: https://jakeseliger.com/2012/05/22/what-you-should-know-befo....

The "deep passion" can also meet academic politics, the need for stable income, the insanity of the peer-reviewed journal system, sadistic or simply indifferent advisors, a lunatic dissertation committee, and so on.

There are likely some solutions to some of these problems; the elimination of tenure could be one, or the creation of the "teaching" PhD that consists of four years of coursework with a relatively small amount of research, with the degree awarded based on whatever papers have been written or contributed to, or whatever experiments have been done.

I think a big part of being "horrendous at being self-directed on long-running, large-scale, multi-step tasks" comes from not being trained on that set of skills. I'm not so sure it's an inherent human trait.

The traditional school system, at least in the US, is almost entirely focused on training obedience and completion of relatively short-term tasks. Homework, tests, maybe the odd paper that's "due in 3 weeks" but really can be done in a mad dash the night before.

Until someone tries to go for a PhD, they may never have needed to plan and execute a large-scale, long-running, poorly-defined task on their own before. School didn't train for that. And they're expected to undo 20+ years of that step-by-step spoon-fed approach, throw that all out the window... and learn a new way of working, mostly on their own, and tackle this huge project. No wonder the failure rate is so high.

> All we ask them to do is write 100 pages or so on a topic of their choosing. They're given four or five years, which comes out to less than a page a week. Yet more than half will fail to complete the task due to attrition. (Yeah, I know this is over-simplified...)

This is extremely simplified.

In my PhD program (engineering), you had to take more credits than a MS student does (and if you did not come in with an MS, you had to take those as well). A total of about 54 credits of just coursework - not research.

Then you have advisors who will not let you graduate unless you have published N papers (and N varies widely) in their preferred journals. And then you write it all for your 100 page thesis. I know all students slow down a lot when they have to write that thesis, but for many students, writing the thesis is the easiest part, and only at the end.

This might be USA vs Europe – where the extreme situation would be no mandatory coursework and publishing no papers at all - the final hundreds of pages work not being a typical 'paper'. Though the PhD itself might be 'only' 3 years long.

> All we ask them to do is write 100 pages or so on a topic of their choosing.

What utter nonsense. You have to do research, that is found good enough and sexy enough to be published to get published. In the meanwhile you have all kinds of other tasks you have to do like teaching or being your professors slave, depending on where you do your PhD. Writing a thesis is part of it, but is just a little part.

I feel like needing structured environments has been increasingly stigmatized as hustle culture and entrepreneurship is increasingly popularized. It's not necessarily wrong but is it the best fit for everyone?

Of course if you are looking at a startup, needing the structured environment is viewed as a weakness, but the successful startup hub itself is a testament to how long a little structure can go.

I am someone who thrives in structure. For example, in academia, I'm not in need of resources, passion for learning, or direction. I simply need an instructor who expects things from me with hard deadlines.

For someone like me who reaches out for advice, all these self improvement methods of hustle and driving your own goals are really non-starters. I know I cannot consistently engage in them. From that lens I will always be a failure, although my work under various organizations is highly valued and well compensated.

The best way for me is to find the best existing structure who goals are aligned with mine, working within it to achieve maximum mutual value, and reassessing whether this structure is still the best for me, or looking elsewhere.

Learn to find structure elsewhere. As someone who struggles with ADHD one of the best things I stumbled across are accountability groups. I have no problem with focus, structure and deadlines when it's something I'm really into - but for everything else it can be a real struggle! Some groups are just like minded people who get together, in my experience the better ones are formalized as part of a paid program.

Probably because a bunch of people that don't do well with unstructured systems probably shouldn't be trying to run structured systems :) So we pay people who are good at organizing to do it.

I set my tasks and deadlines (and the group can help with that if I get stuck), and I have partners in the group that provide that deadline accountability and also the structure of checkins for regular progress. Something I scoffed at first but find to be valuable is just being on a zoom meeting with someone else while working on a task. Not necessarily interacting with them, but just knowing someone else is there helps with focus, tends to keep distractions/procrastination in check, etc.

Probably sounds crazy to most reading this but for those who are wired a bit differently strategies like this can make all the difference.

Usually a PhD requires you advanced science (usually the smallest possible step forward). This requires a lot of the skills you aquired as an undergrad, but what can get you through undergrad (beeing good at exams) is almost useless for advancing science. So we dont know if the people trying to do PhD actually have all the skills needed.

> They're given four or five years,

That's the problem. How many can say of themselves that they are pursuing the same issues with the same esteem as they did two/three years ago?

And as someone who was in academia and has a PhD: You don't write a mere 100 pages in four years. Instead you have to submit tons of papers in between to fulfill the scoring needs of the institution you are affiliated with.

I've spent 3 years in a postgrad program and dropped off good 22 years ago.

I have learned a ton of stuff, some of which was a key part of my day job in 10 years. I've read a number of brilliant papers, and a ton of so-so papers. I had a rather clear idea, which you can currently see mundanely implemented in many IDEs.

But my scale proved to be too grandiose, my understanding of it, too vague, going / flying to conferences scantly reimbursed, and general guidance which might have helped me find a realistic path to completion, mostly non-existent.

The industry, in its better parts, is a much more helpful community, people working both on open-source and commercial projects.

> You're starting with some of the most intelligent, diligent, bright-eyed young people on the planet and letting them work independently in a field that they presumably have a deep passion for.

Most PhDs I know have barely any passion and are just doing it because they don't want to work.

This isn't at all what the modern PhD is like today. They're given four to six years in which there is a pressure to be consistently producing slices of work that no one's seen before.