Comment by fzwang | Hacker News Reader

Comment by fzwang | Hacker News Reader

Parent Full thread

I run a comp sci education program to help students self direct their education[1]. We sometimes reference the OSSU curriculum.

Althought there are lots of benefits to the self-taught route, there are some caveats which students should be aware of. You will have to work harder on the "signaling" and networking. There are definitely social benefits in being associated with a university. And a lack of degree will mean you're "marked"[2], which you'll have to overcome. A setback or mistake will be attributed to your lack of degree, whether justified or not. And some hiring managers can't take the political risk of hiring a non-degreed candidate. Not insurmountable, but this means we work on it from day one. If you do decide to self-direct your education, the benefits are that you learn faster and don't waste time spining the hamster wheel, so to speak, to grind out courses. Everything you learn is in context and relevant. If you realize you miss some fundamentals, you'll just go back and learn those concepts/topics. It's a different way of learning, which imo, is inevitable for technical professions. But it's not for everyone, and some students just vibe with it more.

What's sad is that many students are sort of forced into the self-taught route, because they don't have the financial resources to go to college/university. And if they're not aware of the trade-offs, they could really struggle.

[1] https://www.divepod.to [2] https://en.wikipedia.org/wiki/Markedness

I’ve followed and part-time mentored several people through their self-taught education. There are a lot of pitfalls and traps that can send people down the wrong path if they’re not careful.

One that I did not expect but that seems obvious in retrospect: It’s really easy to start reading Reddit or watching Twitch streams of developers ranting about the industry and think that actual skills don’t matter any more. There’s a temptation to think that you’re a fool to study and practice the job skills because what you really need to do is optimize for interview skills. So they drop everything and starting grinding LeetCode, putting unfinished “side projects” on their GitHub that have all the right things in the README.md (just hope nobody actually looks at the code) and memorizing S.T.A.R. format responses for the common behavioral interview questions.

This strategy actually worked reasonable well for a few years, but the game has changed and most companies are better at catching professional interviewers who don’t know how to do much else.

I should note that this mindset isn’t unique to self-taught people: There’s a parallel epidemic of cheating in college among students who see it as “just a piece of paper” and think they’d be foolish to actually learn the subject material. This also hits hard when they reach graduation and are faced with the current style of interviews which are not as easy as they expected to bluff your way through.

This is something I had to deal with as well. It also surprized me in terms of how limited their information sources are, esp with younger students. One thing I found helpful is to actually introduce them to engineers in person (like a take your kid to work day), which I think grounds them a bit. But this box-checking influence is everywhere, including in the K-12 curriculum. In some ways I understand their perspective. Most schools/teachers do have a box-checking mentality, and I think students intuitively understand that what these "educators" are after is a metric. They don't actually care about real skills. But to your point, the rest of the world actually values competency and it's something students should strive towards for the long-term.

> putting unfinished “side projects” on their GitHub that have all the right things in the README.md (just hope nobody actually looks at the code) and memorizing S.T.A.R. format responses for the common behavioral interview questions.

This perfectly describes my experience when reviewing resumes of grads from certain bootcamps. The program held their hand as evidenced by every student have a similar setup: claiming the cookie cutter 3 month CRUD white labeled webapp as work experience. Everyone on the team is a “co-founder”. Apparently all 4 people "managed a remote team of 4 developers". When you dig into the code, it’s a toy project not intended for any real users. The bulk of their "webapp" is a "case study" page with sections including "the problem", "the solutions", "What is a build a process". It seems these sections were assigned as homework. Their resume includes what things they clicked on in the AWS UI.

In fact, it seems the whole group were instructed to post on HackerNews "who is hiring" with the exact same template. That is the extent of handholding occuring in these bootcamps.

Some challenges as an autodidact:

- Some people assume you lack theoretical/foundational knowledge.

- Guidance/mentorship is harder to come by.

- You’re likely learning on your free time while working. Quality time is hard to carve out.

- It’s harder to get a sense of where you’re at.

- External validation is much more difficult to get. But you need it when you’re searching for a job.

- You inevitably make a lot of decisions that university students don’t have to make. This can be taxing.

- It requires more discipline, because there isn’t anyone checking in or forcing you to demonstrate your learning.

On the other hand, overcoming these challenges is beneficial, especially if you never really stop learning/studying. You pause from time to time, but you pick up again, because there’s always more to learn.

A big advantage that might not be obvious: You pick up niche subjects, simply because they interest you.

You’re not just learning things that seem useful in your context. It’s actually often the other way around: you learn things that you’re curious about and perhaps a year later you encounter a situation that you can solve a problem because of that.

Curiosity is an interesting mechanism. It’s often a better guide at driving your learning than an analytical approach.

I agree. Self-directed education definitely front loads a lot of the problems/decision making early on. I think one of the goals of a good advisor, if you can find one, is to smooth out these issues so they don't become overwhelming.

And to your point about curiosity-driven learning, I'm often shocked at how deep some students get with their learning, if/when they are interested. In some ways, I think we really underestimate how capable young people are. And it's very satisfying, as an educator, to see a student start to embrace continuous learning, as their default mode, and don't interpret learning as a chore or box to check.

Reply on news.ycombinator.com