> Think the programming talent pool is increasingly getting split into run-of-the-mill and superstar developers.

That doesn't conform to my experience (European / working in the industrial sector / hiring developers but not a developer anymore myself) at all.

Pay is mostly based on the sector you work in and the origin of the company you work for. You will have an incredible salary if you work for an American web company especially a big one. You will have a fairly good salary in finance which might become an extremely good salary the closer you get to trading. Meanwhile, some of the guys working with me were experts in sensors fusion, embedded development and highly available systems with decades of experience and were paid peanuts (that was in the defense industry). They all stayed out of a mix of loyalty and the conviction they wouldn't find something as interesting somewhere else (probably right founded from what I have seen).

I wanna upvote this so many times. I used to work in geointelligence surrounded by some of the world's smartest people doing some of the most amazing work I will ever see. We were doing things with pixel data that Google is still at least a decade away from, and the efficiency of the ground processing ingest systems were so insane that when the government tried to give the contract to someone else, it took their software 21+ days to do what we could do in minutes. But it took decades of fine tuning to get to that point, physicists and mathematicians working in close concert with engineers who understood hardware, software, and orbital mechanics equally well.

Meanwhile, I left and am now making double, in some cases triple, what some of my old coworkers are making, all because I'm in a hotter industry infused with VC cash, not because I'm any better.

And I absolutely miss it and understand why some of those people don't leave. The problems you get to work on and the access to classified data and close collaboration with so many world class research laboratories is unmatched. You'll never get that level of intellectual engagement working for a web company until you're in one of the handful of moonshot programs staffed by PhDs, but you will get money, so I chose money.

This does not correlate to my experience working in Los Angeles and Silicon Valley. SV programmers get paid very well, as well as programmers working in L.A. for Google or Facebook. However, among those high paid programmers there were some really high performers and a lot of people who were good but really just like the people who were just getting paid much less somewhere else. One thing was always true, once someone got a really big pay check they were quite convinced that their work was very special.

While the work at your previous employer sounds really cutting edge, how profitable were they?

There also seem to be some areas where highly skilled workers will take a pay cut to work on something they love. This of course leads to lower pay in that field. A common example is the video game industry.

Your previous job sounds like it might have been affected by both lack of a money printer and lots of folks who take lower salary because they love the work

As you might imagine, working on a classified program means you're working for the government, so your company's profitability is inherently a function of the lowest bid mechanism for contracting - to an extent, as we proved they can't just costlessly switch to the cheapest bid unless they want a system that doesn't actually work. It's maybe counterintuitive, as the federal government is the only employer that literally does have the ability to print money, but in practice, even though they often face and solve problems far more complex than any faced by industry, it is often not a money-making endeavor.

Specifically, the point of geointelligence is we're trying to gain a long-term strategic advantage over all other enemy and competitor countries. To the extent we succeed at doing this, the government itself doesn't earn a profit from it. Everyone else does. We try to tax them in order to pay ourselves, but people don't like taxes and they frankly don't understand the ways in which spycraft is benefiting them because all that ever gets publicized are our failures. When what we do works, you never hear about it.

I work at an engineering firm and that is so true. All the schooling and a professional engineering (PE) certification that if a mistake is made with your stamp on it results in you personally not being able to practice for a long time or never again. I see these young, brilliant engineers working on really cool stuff that I also work on, but I am not an engineer. I am self-taught, never finished my physics degree, and went and did whatever job struck my fancy: welder, machinist, IT (been coding since 1978 on Commodore PET 2000), project management, museum art handler, underwater hydraulics and electrical work (senior manager at a water show in SE Asia), ropework (just did a 300 foot drop inspecting a facade, and I am 56 now), entertainment engineering, owned my own design/build shop doing animations, displays, events, stage, TV, etc., 3D work in Prisms (now Houdini), etc... I followed my bliss as J. Campbell would say. Yes, sometimes I was earning way lower than my potential, but I gained knowledge and experience. Besides, I have been making a very good salary starting about 15 years ago. COVID has pushed me to setup a smaller version of the rapid-prototyping, mechatronics, design/build setup I once had, except now on a desktop scale as a side interest, hobby, and potential income earner. I know why the others stayed at your company. I had offers for silly money overseas, but when I saw it was a bunch of networked, low-producing, cocktail-party chameleons who I could not learn from, or enjoy their company, I passed. Life's short. Do what fulfills you and still supports you and your family. Your needs adjust when you are happy!

HN has a very strong web bias, so that comment is likely only looking at that kind of work. Here in Midwestern US, I see the same stratification that you do.

> Think the programming talent pool is increasingly getting split into run-of-the-mill and superstar developers.

Absolutely bullshit also designed to turn workers against each other to the ownership class's benefit.

We live and die by the quality of each other's abstractions. If everyone is writing shit, I can go it alone 1980s nolstagia style, or swim in molasses. If everyone else is writing excellent libraries, I am flying with a jet pack.

There is definitely attempts with certain bootcamps, anti-intellectual languages like Go, to beat programmers into submission. The "superstars" may do fine in the short term, but will not survive this trend long term.

> anti-intellectual languages like Go

Meaning...?

I generally agree with your class argument but don't see how Go fits into it, nor what an "intellectual" language would be.

Perhaps they thought the lack of memory management options in Go is anti-intellectual?

I'm not a huge fan of Go (more of a Python/Rust person--to promote my bias) but I can't think of how it would be "anti-intellectual" (either). The statement was way out of place.

Ignoring prior art, like generic types until 10 years later. Basically it's the evolution of java repeated; the second time is farce.

There are large questions of trade offs, how are we measuring value? Run-of-the-mill works best when things are well understood and stable. Superstar is needed when things get unpredictable and stop working unexpectedly. Think of the large number of JS frameworks which have come and gone over the past decade. Microsoft's .net4 -> core -> .net5

How can we tie compensation to ownership?

We've seen software die as the abstraction falls into disrepair. There's a huge numbers of frameworks to try and solve this, WINE, Dosbox, SCUMMVM, Dolphin.

Flying with a Jetpack is great until it runs out of fuel

It's not working like that; there is no committee deciding your fair pay based on your technical prowess.

It's all down to how well you are able to market yourself. Granted possessing actual competence makes that easier, but really, the world is full of competent, smart but not particularly pushy people who get ripped off.

I agree, the free market decides your wage no only on technical prowess, but also your "influencer score" and your personality traits.

Technically savvy folks with a low influencer score may get paid less.

Technically savvy folks who do not negotiate well may get paid less.

The modern market rewards folks based on their "basked of skills".

I absolutely disagree, the pool of superstar/senior devs is tiny. The vast majority of devs that are past junior are mid-level, don't let the job title fool you

I'd say superstar is probably 10 years minimum of experience. And even then it doesn't just come to a person. Most will forever be mid and that is ok

Seems like you're equating superstar developers with years of experience and that hasn't been my observation. PyTorch was created by an undergrad. I've worked with senior devs that have let engineering pass them by and are now the equivalent of non-programming programmers: https://blog.codinghorror.com/the-nonprogramming-programmer/

I agree that the pool of superstar devs is tiny.

As a data point, in big tech the Sr staff (L7 at G/FB) and higher is no more then 2% if ICs.

For anyone confused, “senior staff” is L7 and rare but “senior” is L5 and usually a majority of SWEs.

Correct. Sr/5 is the expected terminal level for most. Both FB and G expect steady progress to that level or you get fired (if you really can't make it).

L4 is terminal at Google now afaik, but same concept.

Did that change?

I didnt say years of experience is equal to status

But it sure as hell takes years of experience to achieve certain status

It takes building and shipping an outstanding product to achieve certain status, if we’re talking about “superstar“ status. This has almost nothing to do with years of experience. I’d even say most superstar devs achieved it pretty early in their career.

It absolutely does take years to achieve that. Most people that think they are superstar are really just mediocre mis with an ego.

I doubt there are many superstars under the age of 28 minimum

The limited empirical data I've seen shows the idea of superstar programmer in terms of efficiency may be a facade, unless superstar is a 1.5-2x efficiency developer (the slides are concise, there's a corresponding paper published I can't seem to find this morning):

https://resources.sei.cmu.edu/library/asset-view.cfm?assetid...

> For routine tasks, professional programmers have a narrower range of productivity than we first supposed, but almost half of the variation in individual productivity is noise, making programmer rankings suspect.

> Focus only on Programming Effort (limit the scope)

> This is for small programs that should not exceed normal capability

Yes, when you eliminate most of the sources of variation, there is a lot less variation. The most efficient developers aren't more efficient because they churn out boilerplate code faster.

Top talent in big tech are not working as programmers. They work as engineers. They often do a limited amount of actual programming, most of the contribution is leadership, design and architecture.

I agree completely, I think this is really the distinction that is lacking when people talk about superstar developers. Most aren't developing much of anything in terms of raw implementation, they're redesigning existing architectures, theorizing new approaches, bridging domains, etc.

Those skills are often tied to specific domains that are not quickly transferable so what talented engineer A does efficiently or makes them a superstar is specific to engineer A. Talented engineer B probably does something a bit different or has their own sets of expertise they're superstars at.

There are a lot of misguided organizations hiring on the wrong premise though, that there are developers who just churn out implementations at 10-100x speed so they can reduce their workforce. What they should focus on is engineering talent that makes sure they're tackling the correct problems and focus on the right solutions to those problems as opposed to something silly like developer velocity.

You explained the fundamental philosophical divide in software development today. One group sees development as a process of production. They still think of developers as people who produce code. The other group sees software development as a process of problem discovery and definition. Programming is an exploratory learning process that, at its best, refines and condenses human concepts and symbols into concrete representations automated by a computer. In my opinion the latter group will always outperform the former (barring massive resource discrepancies) because they have a more powerful theoretical grip on their activities.

A great design/architect will mean you need to write 1/10th as much code.

Well, not really but maybe 1/2

Dominique Brezinski's latest Spark Summit talk is what cemented my feeling that superstar economics are at play in the current marketplace (in certain fields): https://databricks.com/session_na20/patterns-and-operational...

He talks about the data processing framework he built that's ingesting "hundreds of terabytes, approaching a petabyte a day".

New technologies let small teams do tons of work.

Quite. 10 fresh graduates of the University of North Central Tennessee’s CS program are capable of doing a better job of anything in the CS realm than Notch, Jeff Dean or Linus Torvalds.

I highly doubt that for Dean and Torvalds. They are not programmers, they are software engineers. Both guide others to make 10x or 100x impacts to the world. 10 new grads, in most cases, cannot do that. Let alone this takes for granted both of their algorithms and data structures knowladge.

While Jeff is a great programer, writing code has not been his primary contribution at Google.

When you have multiple people, there's an increasing amount of overhead, from communication, from coordination, needing someone to manage and lead. Now if a single person is able to do the work of multiple people, or even more so multiple people from different disciplines and different teams, then you are gaining much more efficiency. Because you've now replaced 4 people, and then you're moving 4x faster than those 4 people were working together, because there's no overhead required. So now it's actually a 16x speedup. In the same time that it took someone to schedule a meeting with the other team's manager to ask for resources to prioritize time in the next sprint, for someone to review the design and then help out, your superstar already did the whole thing herself after lunch.

A bit off topic, but how does a university student distinguish themselves as a good developer? I have projects I think are interesting on my GitHub and I've worked three internships while in Uni, but how can I make sure I stand out during the job hunt (currently in my last semester) while I'm competing against some of my classmates that didn't know the equality operator could be used outside of an if statement (a real example)?

There is a bit of a marketing element to it as well. You have present yourself as the perfect candidate based on your experience.

Let the employer know how much you want the job. Know the details about the company. Practice answers to questions. Treat this like you’re training to win the US Open. You would be surprised how little effort most people out into preparing.

Also keep in mind that you won’t get every job you apply for and it might not even be your fault. When hiring managers are building a team they may be looking for a specific “personality or subtype” to fill out the team. Where If they had multiple head counts they would probably hire you but are holding out for the perfect fit.

What I did: Post on HN who wants to be hired, work at a small startup for a while, be a driving force between positive change.

My first job was 50k/year and I lost money just by getting to work (gas, fees, parking, etc). Told my boss and in 6 months I got a >50% raise to 80k/year + stocks + parking space.

Next job was 120k + stocks + remote for a company in the valley.

Each job hop will net you a pay increase similar to this. It'll also give you new perspective on software engineering.

My email is in my profile if you want to discuss in more detail!

"Game recognizes game."

People who are really good can recognize each other.

At some point in the interview process they'll sit you down in the room with a couple of their other really good people, you'll "grok" each other and be hired. (If that never happens, decline the offer, they have no one and no clue.)

No-one expects a fresh college grad to be a rockstar dev. The few places that do, probably have highly specialized recruitment pipelines, and look for people with the "correct" features and milestones, like monitoring people that have participated in math / informatics olympiads, that have gone to the "correct" schools, and have been mentored by important figures.

Those places are rare. But for the majority, my tip would be:

- Get comfortable with a couple of languages, learn good style, and get familiar with the toolkit.

- Learn the fundamentals of CS.

- Commit to some side projects.

- Seek out internships/work experience

- Do fairly well in school.

- Learn teamwork and soft skills.

I don't think any single/individual point above is enough to signal "awesome developer", but the sum of them will surely correlate with high potential.

I work for a much smaller company now than I did a few years ago, so I'm interviewing at a much lower rate. That said, if you have a single project on Github and can discusss it intelligently, you are already head and shoulders above the vast majority of people I've interviewed over the last 20 years.

You need to demonstrate that you understand and can modify code in real project. Many programmers are code illiterates, job interview is all about filtering them out.

Small trivial projects on github profile do not count for much. Contributions to some established OS project are great.

I stood out by taking part in my uni's Formula Student group and wrote safety-critical code for the car. This enabled me to find an interesting job straight out of uni where my code would be deployed to millions of cars around the world.

Yes, I feel pay for developers is going to become more bimodal as time goes on; similar to lawyer compensation.

Yep, agreed. Here's an article for the bimodal nature of starting salaries for lawyers for interested parties: https://www.biglawinvestor.com/bimodal-salary-distribution-c...

The upper end of the biglaw salaries was capped because of an unspoken "cartel pricing cap". Big law firms don't want to compete for the upper-echelon talent.

Not sure if tech companies will be able to informally collude in such a manner to cap the upper end of the bimodal curve.

They have before[1]. I think one reason salaries started to climb so much recently in silicon valley was because this collusion to suppress salaries ended.

[1]https://www.cnet.com/news/apple-google-others-settle-anti-po...

BigLaw is getting squeezed by its clients. There are increasing numbers of companies that don’t want to pay hundreds of dollars an hour for fresh grads no matter how bright. That leaves law firms with the choice of either subsidizing young associates, which is dangerous given the competition for top partners, or finding some way to cut pay. While few have wanted to break away from marquee associate pay, the quiet rise of staff attorneys at BigLaw represents the breakdown of the bimodal system.

I know a relatively new lawyer that moved from a top law firm in NYC to a tech firm on the west coast and got a huge increase in pay, especially per hour worked.

Getting a rise per hour worked in law by leaving the big law firms is often "easy". My ex worked for one of the top lawfirms in London, and quickly realised their very high headline pay (straight out of uni into ~2x median UK pay, and increasing at far above average rates for 10+ years) gave the new graduates lower per-hour pay than their secretaries.

It was basically a death-march to see who'd stick it out long enough. Only a few would make partner anyway, so they needed to shed a lot of people to make their cost structure work, and so they'd work them into the ground and use who stays as the selection method at the lower end.

In case anyone want to read more about this, that employment structure is called the “tournament model”.

It was a raise by absolute value also, not just per hour. Added bonus of no income tax in Washington.

This is fascinating. I wonder if doctors and engineers and other knowledge-based professions have a similar distribution if you look hard enough.