Excellence is what matters. It can be found in universities (and is in theory what universities are for), but in my professional experience, a C.S. degree is no guarantee that one has been trained with excellent taste and judgement. In fact, I usually react to the new hires with C.S. degrees with some degree of frustration at what schools are teaching these days.

Excellence can also be found in independent reading, and it can be learned with professional mentors as well.

You can learn it from AI, too, actually. I'm skeptical about the central use of AI in writing code directly, but find it incredible as a teaching tool. It is especially impressive when I use it as something between a tutor and a search engine, talking to it a lot about the best way to do things and asking it about standard methods and resources. It's not infallible, to be sure, but it's this incredible combination of tutor and textbook, with a better connection to the zeitgeist than all but the very best of tutors and textbooks.

In this way, AI is not dissimilar from school, independent reading, or work experience: you can use these tools in different ways, to serve different goals. If you focus on credential, skill, productivity, and money, you may entirely miss out on excellence in your craft. On the other hand, you can pursue excellence in any of these contexts.

Classic gatekeeping argument insisting a CS degree is necessary to write good software. I don’t have one and have 20+ year successful career with accolades, know many others in the same boat. Obviously there are CS degree people who were also successful, and the flip side comes for both camps, I’ve worked with absolutely awful engineers with or without a CS degree.

It is a signal, not the signal.

Guy writes like he’s the main character when he’s probably dev #48373 at CrudCorp. I bet the internal company tool he works on is pretty slick though.

Most smart people I know don’t like toy leetcode problems, they like real problems which are orders of magnitude harder and not solvable by LLMs. Leetcode is for people who got good grades and want to feel smart by learning the rote tricks to solve leetcode.

I'm not sure how this kind of software gatekeeping squares with our new glorious future where everyone is happily running bespoke vibecoded applications tailored specifically to their needs.

I bristled a bit at the scalpel "you don't want to give this tool to the wrong person" analogy. While I agree that there are certain classes of software (like banking/finance, healthcare, etc) that you don't want just anyone writing, but you don't need to have a C.S. degree to whip up a SwiftUI Music.app replacement so you can listen to dad Rock while you work in the basement.

I have a CS degree. I learned some useful things. However, I graduated in 1998. I've had to learn a bunch of stuff since then because it didn't exist in 1998.

For folks that don't have CS degrees, they just have a little more to learn outside college. When I started my career, a significant majority of my colleagues didn't have CS degrees. It just wasn't a popular major in the 90s.

Among the best developers I've ever worked with never went to college at all.

3 positives from a degree are:

1. the complete demystification of computing. If you’re writing c or assembly and working on OS concepts there’s a sort of aha moment that the computer is not a mystery.

2. The degree pushes you past the hump where a lot of folks get stuck on when learning to program. There’s a seeming spot where things makes sense and you actually understand what’s going on. Forcing yourself to stick with it for 4 years where you’re taught by experts is helpful for that.

3. It gives you a space to explore a wide array of topics. Whether they are useful to an employer or not.

I assume folks can accomplish these 3 on their own. But it would be more or less building your own balanced curriculum and also trusting as you start that your curriculum will work. Whereas within the school system the trust is baked in: if you successfully complete these assignments you will learn foundational concepts.

While I have a CS degree from a good university, I think I see things differently than the author here. I think CS degrees are great overviews of the field and give you 4 years to practice programming basics without needing to worry about your job. You also might meet some smart classmates and learn from them too.

However, I think the core issue that individuals run into is that the industry does not stop to train engineers in any kind of formal way. If you do not address this, then your skills will never adequately develop over time. This goes for CS degree holders too, they just might have some more confidence initially from the 4 years of marinating in the material.

There are so many ways to develop your skills:

- cs degree, sure

- pair programming with people that have the skills you want

- helping others with their problems

- building personal tools

- reading blogs, watching videos, and generally soaking up external opinions

- taking projects slightly outside of your skill level

- self-reflection and writing down the problems you have, then actively seeking help with each specific problem

- learning to read error messages so you don't sit there slackjawed when red text shows up

- learning to read documentation and scan for useful APIs

- reading books, if that's a method that works for you

- doing retreats like Recurse Center that push you

- SRS practice with tools like Execute Program

- paid courses with work edu budgets

No one will really make you do these things though, so it is very sink or swim. You don't have to do all of these things, but you likely need to be on some kind of upward trajectory, particularly for the first solid chunk of your career. For me, this is mostly work I would do as part of the job, during 9-5 work hours.

I think the post has kinda contradictory parts. The beginning says that a requirement of formal training, like in medicine or architecture, or transport planning, is desirable.

But then a number of examples are given, that suggest that degrees are orthogonal to skills and promotion, and instead being aggressive promotes better than excellency.

So how a compulsory degree would fix this corrupted promotion mechanism?

A lot of the leet code problems are "math shaped". LLM's are pretty strong at that.

My guess is that interview questions would test your skills combined with the llm, because when testing skill without, the difference is just too big.

I got into this stuff super late. Have kids and a job. Would love nothing more than to go get a proper cs degree. Would probably go for computer engineering. Alas, my undergrad GPA was poor, so grad school is out. Can't picture dropping six digits on a degree from like Trump University or wherever, and seems silly if not impossible to go get a second undergrad. Sucks because my alma mater had an amazing cs dept, I just didn't know or care back then. Oh well.

Especially with advent of agents, I have dropped all illusions of ever getting paid for this outside of like freelance work. Will keep working through cs:app3e for the love of the game.

100% agree. The second I got a degree, gatekeeping stopped.

Becoming an MBA candidate, the barriers were minimized even more, if not eliminated.

I can only imagine if it was a top ten school instead of top 30.

Combine a top 10 with a proper computer science degree. If only I knew and understood this was the “proper” path a decade ago or more.

Taking the harder path helped me become who I am today, but, it sure is easier to take the easier path.

If you are looking for a practical computer science degree, Georgia Tech’s OMSCS feels tailor made for late 20s and 30s. There’s no age restriction of course, but if you are younger, on campus networking is better.

I plan to do it sometime after my MBA.

> Experience has shown that, things rarely work out in the long-term if you aim to have a successful career without a degree or by skirting coding interviews. Lacking these skills in the current conditions might be career-ending.

Perhaps your experience is unique and not at all representative. This reads very out touch.

Edit. Here's some data.. none of which indicates "rarely works out".

SO survey 2025:

20% of respondants who were professionally coding did not have a 4 year degree. Only 17% of all respondants said they learned to code from school.

SO survey 2017:

> The remaining 20.9% said they had majored in other fields such as business, the social sciences, natural sciences, non-computer engineering, or the arts.

> Of current professional developers, 32% said their formal education was not very important or not important at all to their career success. This is not entirely surprising given that 90% of developers overall consider themselves at least somewhat self-taught: a formal degree is only one aspect of their education, and so much of their practical day-to-day work depends on their company’s individual tech stack decisions.

This is a very dated view. We're in the process of offloading that to the machines.

Based on the comments here and the content of the post... I sort of agree. But maybe less so directly with their gatekeeping. But maybe we should formalize closer to "real professions".

And if we think college degrees don't matter, there's plenty of professions that require time on the job, rotations, flying low/no passenger planes, and so on.

We require people get a license to own a firearm (or ostensibly should...). And AI can be an extremely powerful weapon.

I see a gatekeeping argument. I don't agree with their requirements. I'd rather see that regulation than banning models.

It also has the side effect of preventing AI from taking jobs as aggressively, keeping it at bay in schools, and getting people to maybe respect(?) / acknowledge the dangers involved here.

So says "zaksa".

Follow up question. Who is zaksa?

> They either focus on the wrong things when developing software, they have a very shallow knowledge and lack of regard of how difficult good engineering is, or they’re utterly clueless and rely on good soft skills and politics.

I’ve worked with plenty of people like this that had degrees. A few of them even have masters.

> The people I wrote above weren’t considered bad performers.

Did management - the people who are supposed to deal with the issues described - have degrees?

Even if you take the problem seriously as he describes it, it is delusional to imagine that a formal degree would be the solution.

Not that I would consider a "licensure required to use a sharp knife" kind of world, even if it would solve some problems.

Crazy. It depends on the kind of software being built. Working on embedded systems for avionics is an entirely different animal than working CRM software development as an example. Even Wozniak didn’t have a degree until 1986, long after Apple went public.

Most of the people I worked with at Apple didn’t have CS degrees yet they built the thing many of you are using to read this.

Much of software development is deciding what to build rather than how to build it — that takes skills far beyond CS. We are building software for people to use rather than building it for its own sake.

CS people might know algorithms but do they understand why they even need it? How does it help the user? Some of the worst places I’ve ever worked had lots of CS degree people who spent more time complaining about framework defaults than building something people are able to quickly use and pay money for.

Inventors of Perl, C, Turbo Pascal, Typescript, Delphi, C#, FORTRAN, PHP, and COBOL pioneer Grace Hopper — no CS degrees.

Jack Dorsey, Peter Thiel, Zuckerberg, Chris Wanstrath (GitHub,) Arash Ferdowsi (Dropbox.) — no CS degree.

Stewart Butterfield (Slack) had degrees in philosophy! Reid Hoffman — also a liberal arts guy.

Avie Tevanian (Architect of MacOS,) — no CS degree. And while Steve Jobs wasn’t as technically astute as Wozniak — he still did tech and of course, no degree at all.

I could go on all day — I only mentioned the famous ones. At Apple I could list a hundred top engineers with no CS degree. I am sure most people here could.

The CS degree gatekeeping is complete ridiculousness.

I am biased though— I spent 7 years at Apple and have been in this business 16 years and my degree? Journalism.

CS people might be able to answer questions, but many times they don’t know the questions to ask. Tech has room for everyone and everyone brings a different skill set to the party. Not all doctors need to be neurosurgeons.

i dont think a whole degree is required for software dev, prolly better to just make it a major in some other degree like eng or sci

Milk takes 3 days to age. This is already spoiled.

Yeah, no.

[dead]

While I agree with this in principle, I think the market is a separate problem. As much as this is correct, the market is not going to select for competence. The market is going to select for affordability.

Ultimately the most affordable worker of the future is going to be the person who knows just enough to prompt an agent to first build a feature and second fix it "fast enough" when it ultimately breaks.

Companies might have a couple of people who actually know how to design systems and think deeply on hand for overall architecture, load balancing design, and serious problems but this will be a small cadre of employees. In some sense, I suspect, in relation to degrees, computer science will eventually look like the field of curation: an extremely small pool of outstanding positions looking to be filled by an orders of magnitude greater number of degree holders.

Unless the bubble does eventually burst, in which case people will hopefully return to their senses.

Well written. I have a similar background and I agree with most of your observations and conclusions. Not the part about prestigious universities catering to 'rich kids", though, at least not any STEM departments.

I removed dot zip from the interwebs using dnsmasq. https://web.archive.org/web/20260822143156/https://zaksa.zip...

address=/.app/0.0.0.0

address=/.vercel.app/0.0.0.0

address=/.zip/0.0.0.0

address=/.mov/0.0.0.0

address=/.so/0.0.0.0