Comment by effingwewt | Hacker News Reader

Comment by effingwewt | Hacker News Reader

Parent Full thread

History rhymes.

This whole thing was a repeat of the dotcom bubble. Back then it was 'learn web design!'

Here too we had bootcamps to learn to code in weeks. At that point the bubble was reaching critical mass.

Same damn thing happened before.

I tried warning people- here on HN the sentiment was 'we are indisposable'.

When the cost-cutting measures and hiring freezes began I tried warning again and was told 'its not a hiring freeze at Uber, we are fine...just stopped hiring for certain departments (definition of a hiring freeze)'.

Then Twitter began the firing dominoes and here we are.

Thing is it was their company culture and management that made this happen.

Unreal perks, infantilization of the workforce, telling them how important they are, how they are changing the world, they are the 10x programmer.

I can't believe our workforce is going through this again only 20 years later. Seems housing is headed the same way.

We need protections for people, not companies. This is going to keeo happening unless we stop it.

I agree with you it’s in a lot of way similar to the dotcom bubble but the issue seems to me simpler than you make it seems. Propped by an easy access to cheap investment fonds, companies massively over hired developers to work on projets which have actually no value. The massive demand leads to under qualified people being hired. The current layoff is a return to a sustainable situation. And even then if you look at the number for the entire market we are not even back to a prepandemic number.

In control systems, a common tradeoff is between a fast response that overshoots or a slow response that does not overshoot.

We've seen that pace of innovation matters and if your goal is to maximize the pace of innovation, you want that fast response even if it means overshooting somewhat whenever there's a step change.

I am not sure if you mean that dirt-cheap money accelerated the pace of innovation.

But, if that is what you mean, i.e. historically low interest rates accelerate the pace of innovation, we would witness innovation gains in countries with negative or zero interest rates. But we haven't seen any shift.

There's a difference between malinvestment and investing in projects that don't go anywhere.

This is the difference between burning cash in a bonfire, vs using $1 dollar to make $0.XX cents.

I'd argue the system was rife with the former, not the latter

People really can't predict what investments represent the bonfire vs. good investments, and even if they get that right, it's even harder to get the timing right.

For example myspace vs. facebook, webvan vs. amazon vs. instacart. You can also look at physical industries like the aerospace boom after WW2 that brought us into the jet age. The high influx of cash allows huge numbers of companies to bloom, representing the testing of lots of ideas and designs in parallel. Many aerospace companies failed, just as many tech companies have failed, but those tests accelerate innovation.

“We've seen that pace of innovation matters and if your goal is to maximize the pace of innovation, you want that fast response even if it means overshooting somewhat whenever there's a step change.”

Does the pace of innovation really matter as much as you think? Or is it possible execution on the business side can matter more?

Also, it has been proven that adding more software developers can actually slow down a project. Increased hiring, therefore, does not necessarily correlate with faster innovation.

> companies massively over hired developers to work on projects which have actually no value.

It's still unclear to me the layoffs are affecting actual engineers, outside of Meta and Twitter.

Meta is trying to execute one of the biggest company pivot in tech's history after completely giving up on the metaverse. Industry-wide they account for a huge number of laid-off developers.

In the meantime, Twitter's layoffs are minimal (it's always been a small company) but extremely visible due to the company's new owner's very public persona.

Elsewhere, it's pretty close to the normal number of layoffs large tech companies do every year. It's a little known fact to outsiders, but tech companies trim underperformers every year from their workforces. The media almost never report on it, except this year they did because Twitter and Meta were in the news cycle.

These stats also have "Software Engineers" as a category alongside "QA Testers", "Web Devs" and "Application Engineers" listed as different jobs, so I'm skeptical of how accurate they are.

> after completely giving up on the metaverse

They have? What are they up to now? I guess I haven't been keeping up!

AI. [0]

Meta is a huge company. Switching to AI and spinning down their VR investments means a large restructuring.

[0] https://qz.com/meta-layoffs-2023-jobs-metaverse-ai-185019657...

This is surely a gross oversimplification of what's happening. If you see layoffs.fyi, there are a lot of rows with more than 5% fired or where the numbers are in several hundreds to thousands.

While some companies are known to regularly use the tactics you mention (and a lot others also do the same from time to time), they don't use mass firing as a means.

> layoffs.fyi

Website started tracking sometime in 2022.

> there are a lot of rows with more than 5% fired or where the numbers are in several hundreds to thousands.

These include "Non-Technical staff" which have been the majority of laid off employees. [0] [1] [2] [3] [4]

> they don't use mass firing as a means.

They do. Google, Amazon, Microsoft and Apple all had yearly layoffs for underperformers, while maintaining growing headcounts. Target was around 5-10%. The difference is these layoffs were not nearly as talked about in the media.

[0] https://interviewing.io/blog/2022-layoffs-engineers-vs-other...

[1] https://www.bloomberg.com/news/articles/2023-01-24/tech-layo...

[2] https://www.computerworld.com/article/3690309/about-those-te...

[3] https://www.gartner.com/en/newsroom/press-releases/2023-03-0...

[4] https://techreport.com/news/3493451/microsoft-layoffs-ethics...

A lot of the people who got laid off from Google were not "under performers".

> I tried warning people- here on HN the sentiment was 'we are indisposable'.

Maybe if you were actually any good. I reckon at least 60% of these bootcamp alumni were just dead weight. I've interviewed so many people who "learned to code" but had no meaningful logical reasoning or problem solving skills. They could maybe understand syntax, but for the most part they just have some basic familiarity with various frameworks and tools. They don't know how any of it actually works.

Experienced developers are often mediocre communicators, so it's not even like they'd be able to explain what needs to be done clearly enough for people without problem solving skills of their own to run with the ball once its been handed to them. Developers are there for their programming skills, not pedagogical ones, so it's just not a realistic staffing model to be this way.

I had friends that did bootcamps and this mirrors my experience. They know how to use Express a little bit, but not how the middleware works or what HTTP status codes mean.

They can write code if I translate it to requirements and show them something similar already in the code. They tend to flail around when handed a more open ended problem like "we need an endpoint that will return the last 5 purchases made by a particular user". They don't yet grok the kind of questions to ask like where should the data come from and what format should I return it in?

I don't mean to disparage them as students, I just think they were sold a lie. New programmers probably can learn the very basics of a language and framework in 8 weeks, but learning to solve open-ended problems takes a fair bit of practice for most people. There isn't really a shortcut through that.

Yeah your choice of the word "practice" is the important part. A lot of people imagine that developing skills is like plugging into the Matrix and downloading knowledge into your brain. But it's about cultivating a style and approach to thinking that only works through practice.

The first year of law school is often said to be the hardest for the same reason, they're breaking down your intuitions and teaching you to think like a lawyer. I went through the same process as a social scientist. Every discipline has this acculturation process where its disciples are trained in the skills and values of the discipline. It's not just information transfer. You can study sheet music all day, but unless you actually PLAY you will never get it. It's not something you can just boot-camp in a few months.

If people have the skills or values they developed from other disciplines that they can transfer over then the boot-camp can work. If you learned how to do logical reasoning from being a philosophy major, then you'd probably figure out symbolic logic and learn to program pretty easily. If you figured out how systems are put together from a career as an electrician or mechanic, then you can probably apply those skills to architect software too. But you need some foundation there to build on and I don't think a lot of people just getting into it because they heard it's a lucrative career have that.

> or what HTTP status codes mean.

A junior hire I worked with a little while ago had actually done a college class on web development. I must admit the more senior engineers did rib him a little about it, being a little incredulous that a serious college would teach web dev and give out college credits for it. But nowadays there's even one at Stanford so we were wrong.

Anyways, the first thing they learned in that class was the HTTP protocol.

> I just think they were sold a lie

At the same time, where are the aerospace engineering bootcamps? Where's the law or medicine bootcamp? Did the people enrolling really thing they would be on-par with actual engineers working in the field after 12 weeks?

I totally think colleges could teach HTTP. There's plenty of time to do it in. I generally think colleges do a meh to poor job of preparing CS grads for the real world, but they're leaps and bounds better than bootcamps. I'd probably give 50/50 odds on college grads knowing (or at least being able to remember) HTTP status codes.

> At the same time, where are the aerospace engineering bootcamps? Where's the law or medicine bootcamp? Did the people enrolling really thing they would be on-par with actual engineers working in the field after 12 weeks?

2 thoughts on that.

The first is that at least 2 of those 3 (not sure about aerospace engineers) require a specific degree (in most states). It would be a bold faced lie to tell anyone that the AMA is going to recognize them after a bootcamp, and certainly punishable under false advertisement laws. They're probably right that someone could do a bootcamp and be on-par with actual engineers. It might have to be a very specific someone with a very specific background, but it's probably possible.

The second is that yes, I think people do think that. I've had a couple of friends that have done them, and they were convinced that they could pay $12k for a 9 week bootcamp and be a desirable and productive developer. I tried to convince them otherwise, but failed. I hate them because I feel like they're selling an expensive, unrealistic dream to people who have no way to know better.

> I totally think colleges could teach HTTP.

They do. Anyone who takes a web dev class or takes a Senior Project Course should be familiar with it. I was surprised bootcamp grads didn't know about it since it's most often the first slide of a web-dev class.

> I generally think colleges do a meh to poor job of preparing CS grads for the real world,

Often, an issue I've observed is that students will minor in CS and major in something else, or the way the CS program is structured, you can major in CS with mostly math classes. Which is the way to go if a student is aiming at grad school, but not the best in industry.

> They're probably right that someone could do a bootcamp and be on-par with actual engineers.

I'm very skeptical of that claim. I've never seen it play out. Cramming a 4 years degree into 12 weeks?

> I hate them because I feel like they're selling an expensive, unrealistic dream to people who have no way to know better.

I've interviewed someone from a bootcamp who, after 50 minutes, could not come up with a function to count the words in a string. I use it as a warmup question for freshmen who have completed one semester or CS and it usually takes them about 10 minutes.

So yeah, I completely agree if someone paid 12k for that, they got scammed.

> had no meaningful logical reasoning or problem solving skills

This is 70% of software development, this is why I think people from the hard sciences often make better software engineers (in the true engineering sense of the word) than people actually trained in CS. Most CS folks can't experiment their way out of a paper bag.

Probably because CS now has become synonymous with CRUD web development.

I am talking about BigTech top 10 CS school grads. They literally don't understand the scientific method or how to use mathematics to design their systems.

I take huge exception with this remark. I graduated with a BS and MS engineering at a top-20 US engineering school. I also earned an MS at a top-25 computer science program. I found the CS program dramatically more challenging. My CS classmates were some of the brightest problem solvers I've ever worked with in my life.

Then in my career, I have worked alongside some brilliant software developers who had academic backgrounds in non-technical fields like history or English.

I don't know if there's a way to assess someone's problem solving ability based on their field of academic study. It would seem that those from science and engineering fields (CS inclusive) would generally fare better.

Your dismissive comment about "most CS folks" is just offensive.

You are a data point of one, and I have no idea how well you can apply the scientific method to solve problems and understand the state of your system.

I am not judging you, my statement was a generalization.

I would also say, most people in the hard sciences that are not CS cannot code their way out of a paper bag.

I agree with you, some of the best, most rigorous thinkers I have come across had a liberal arts background.

This isn't a quantitative statement about the kinds of people, but the things they focus on. I think CS is hard, but one thing that seems to be repeated is that super capable people can get by operating in an unscientific open-loop fashion which causes cognitive blind spots, and those blind spots are in using experimentation and the scientific method to solve problems.

> This is 70% of software development

Which is why I refuse to use the term "Software Engineer."

Sorry, Java CRUD and copy-pasted React snippets from SO are the furthest thing from an engineering discipline. MOST jobs are web development, even if you're a "backend" developer.

There probably is some legitimate engineering in embedded or development that requires an authentic scientific and engineering background, but considering the majority of code written is CRUD stuff...that ain't it.

> There probably is some legitimate engineering in embedded or development that requires an authentic scientific and engineering background, but considering the majority of code written is CRUD stuff...that ain't it.

I suppose the OS, database, networking and infrastructure to run those CRUD apps as well as the languages and runtimes magically appeared out of nowhere.

> Developers are there for their programming skills, not pedagogical ones

This mindset will limit your advancement, teaching and mentorship work becomes increasingly key for senior and senior+ roles.

But I suppose by the time someone has a senior title and 5+ years of experience they’re not really in the category being discussed in this article.

> This mindset will limit your advancement, teaching and mentorship work becomes increasingly key for senior and senior+ roles.

Yes but there aren't enough of them, and they're not being paid to teach fundamentals to fresh faced entry-level hires. We'd expect them to teach higher level things than the basics of how to program. The skill set to be able to teach and explain to a complete naif is a little different from the skill set to communicate with someone who has already been inducted into the jargon and practices of your field.

Reply on news.ycombinator.com