disclaimer:

This article has a direct intention to offend or upset someone. It was written in a fit of anger and serves as a logical response to what's currently happening in the market. Nevertheless, the author respects all confessions, genders, as well as all real developers and people who use AI as an assistant and this is basically humor anyway. If you don't like it, then you didn't get it. Peace ☮

TL;DR The main theme of my internal dialogue that turned into an article: Are refugees bad because they live poorly, or is it a bad country that failed to integrate them? Or maybe the cause lies in the place that produced these refugees in the first place? Let's figure out how this black mirror reflected the IT market.

intro

In this brief article, I want to focus specifically on vibe coders, the so-called wolves and their impact on the market. I'll discuss what the industry is experiencing right now, potential outcomes, and what ordinary developers should do about it. This piece serves as a direct follow-up to my previous article about wolves in IT from last year, which will be removed as it's no longer relevant. And the key question: how to recognize AI-generated code or articles, and how to spot them in the future.

wolves

This concept is more widespread on the Eurasian continent, especially in CIS countries and parts of Asia. Though this term is slowly migrating into Western culture. Wolves, jackals, dogs. I think it's intuitively clear we're talking about far from the most honest people. So who are these guys?

Well then. We're talking about the most ordinary and long-familiar SCAM. Yes, these are regular fraudsters. The movement gained huge popularity in Russia, Kazakhstan, Belarus, Ukraine, Uzbekistan, and a couple other countries. Mainly thanks to certain bloggers. We'll break down the causes and origins a bit later, but right now let's talk about their mindset.

Yeah, the movement itself been around forever. These are ordinary crooks who make their living by pretending to be other people. This is (and remains) especially popular in India and Pakistan, apparently due to their social inclination toward this kind of activity. But it was the CIS that gave this line of work its name. However, unlike the Microsoft Tech Support guys, the Wolves took a more unique path and actually started doing something. Well... almost, and not all of them.

history of emergence

Promises of a Klondike, perfect working conditions, and even water coolers in the office were successfully sold to ordinary folks in countries where the average salary in the early 2010s was around $500. The chance to work from home 3 hours a day for one Bugatti per month motivated huge numbers of students to rush headfirst into universities to get that coveted offer. And lo and behold, some actually succeeded. At this point, people with genuine interest in engineering started mixing with simple hamsters who smelled profit. And... some of these hamsters realized it wasn't that easy.

renaissance

Skipping the ode about the inadequacy of most universities and other bullshit, let's talk straight about how many guys couldn't finish this path. At that time, the IT market bubble wasn't so inflated and even with minimal knowledge you could land that coveted junior HTML/CSS engineer and windowsill-sweeping position. Hell, even interviews back then looked like kindergarten dialogue - "you got a Lamborghini, right? Well then we're friends!" Nobody really checked knowledge quality and many still had a chance to enter the market.

And even then, people started appearing who figured out the trick. They mimicked, got positions through friend recommendations, and even worked for free to prove their worth.

But, as it happens, years went by, technology developed, the bubble inflated, and these prehistoric methods lost relevance. At this point, courses entered the market. Bootcamps, damn it.

post-renaissance

Bootcamp leeches, parasites. They started manipulating the dreams of trusting high schoolers, forcing them to take out loans for training. They have a pretty complex scheme for interacting with the market, including covering Big Tech's orders for temporary "meat," closing contracts, justifying investors, and other stuff. Basically, these guys were profiting from both sides - huh, smart enough.

However, from the average person's perspective, they got shitty education from what was essentially yesterday's student, were left with debt, and often never found a job.

And then SCAM BOOTCAMPS appeared - BOOO! Scary? No? Well, it fucking scares me.

downfall

The aforementioned events briefly outlined led to the logical creation of a structure that spawned the movement and subsequently organizations that began pretending, and then teaching others how to pretend, in order to land that coveted position. Yes, the Wolves' main goal is to bullshit the employer: the HR recruiter, the interviewing tech guy, and even the team lead. Answer all questions so that you appear to be a harmless smart sheep. Sometimes, such people actually break through a bunch of psychological problems (if they're not psychopaths, of course) and become who they were pretending to be - and this is truly the most positive and rarest outcome.

Most often their cycle is simple:

Yeah, these humanities majors will do anything to dodge calculus...

consequences

I confess! Our team accidentally let one of these wolves slip through. He managed to fool even seniors whose skills I'm 100% confident in. They have many methods, by the way. And especially now, with all these AI tools, it's way easier than cheating on an exam. Meanwhile, real engineers don't do this and often spend a year job hunting, while wolves are sitting pretty with 4 jobs.

Sadly, tthe market encourages this. They dump tasks on the same Indians or hardworking Chinese who are ready to do their work for 10 times cheaper. And what's even worse, they can just work with AI. Imagine you hired an AI agent for $280k/year.

I don't feel sorry for Big Tech and employers. The broken hiring system, tons of interview stages, inflated requirements gave birth to this monster. The market created them itself, not like these sons of bitches decided to screw us over out of nowhere. A man-made golem that destroyed its creator.

solution

There isn't one. Really. The market will find a solution itself, sooner or later. A million mistakes, billions in lost money, heaps of teamleads gone gray with rage.

For now, here's my little list of how to spot a wolf in sheep's clothing based on our experience:

1. They talk a lot. Lots of details, lots of specifics, lots of minutiae. In general, that's how you can spot any liar.
2. Live coding or whatever won't save you - they've already learned, they can already deceive at this stage too. Same with dumb technical questions.
3. Someone vouched for the person. Networking! Yes, if there's a trusted contact, especially someone who has no benefit from hiring said engineer, that already adds credibility.
4. Probationary period. Yes, that's exactly what it was invented for. Just overwhelm the person with tasks, monitor them, invite them to all calls, chat outside of work. During this time you have a chance to catch the bastard red-handed.

The rest doesn't work, or will stop working soon. Yeah, we've essentially devolved into good old feudalism, where one person has to vouch for another person. The online hiring market is dead and we killed it. Well, not we - more like you guys, I never killed anyone, just interviewed sometimes.

And here we should smoothly transition to the topic of vibecoders! SMOOTH TRANSITION WHOOOOSH. Basically it correlates very closely with wolves blah blah, let's get to the point.

vibecoders

This is a pretty loud topic that I won't talk about for long, and you can type this word into YouTube and watch one of the million videos that came out this past week. In short, these are people who chose not to program like engineers, but to program as their heart tells them to.

The main thing to remember is the necessity of human hands and the human factor in development overall, especially if you're developing something more complex than a web page without a backend for the local bakery. And even there, it's important to follow clean code rules, however ethereal that might sound, and review everything manually.

Yes, I perfectly understand that artificial intelligence will eventually learn and grow into something more capable than it is at this stage, and that human influence in this field will decline with each passing day and each news update. However, this doesn't remove humans from the system entirely. We're just going back to a system where all the offices of worker bees in terms of IT were managed by just one anykey dude.

That's how it will be in 99 percent of cases, by my assumptions, in all future startups and small businesses where IT infrastructure will be handled by one technical specialist who understands a little bit of everything on a surface level, just enough to realize where there are errors, where there are inaccuracies, mistakes, and questionable moments. Because artificial intelligence overall is already addressing the vulnerability stage and will learn to eliminate them in a timely manner in the future.

As of writing this article on September 1st, 2025, absolutely all of my colleagues use AI agents as an assistant - Claude, ChatGPT, etc. At the moment, this is truly a useful thing that allows you to cut down on a huge amount of your time, letting you spend it on truly more important things that, again, at this moment artificial intelligence cannot take on. It's just a matter of time. And each day the number of people in companies decreases because most of them were simply doing routine work. Architecture decisions, management, leadership, and high-level technical corrections were made and are still made by a select few.

Based on this, I currently don't consider using an AI assistant as something bad simply because that would be idiocy of the highest degree. However, in an official position at the moment, vibeoding is unsustainable as a concept and is merely a toy that will later grow into something truly worthwhile, but we need to live to see that moment.

I'll be honest, I'm tired from typing this article all night, and even now I'm just recording a voice message to give my hands a rest, feeding it to Claude, who nevertheless I still need to check, refactor, insert into MDX, and handle some things that I don't trust agents with, but it's just a matter of time.

consicuenses

At the moment, the market is in a stage of stagnation, if not sharp decline - the bubble burst in pre-COVID times. A huge number of people have been in a buffer zone all this time, not understanding the scale of what happened. Essentially, the emergence of scam bootcamps itself marked the end of the current system of hiring, training, and development. And internal company structures.

I believe that due to the huge leap in technology, naturally a certain downgrade will follow - meaning we'll roll back over time to the previous mode when a department consisted not of a thousand, but of ten people simply doing their job competently. Back then they actually did this, but at the moment, as I said, staff was forced to grow to 1000 to cover all the demands.

Yes, we really will return to that bright past and I sincerely believe in this and dream about it every night before bed.

That companies will only employ people who are genuinely interested in doing their work, not those who do it only for money. And in this regard, artificial intelligence really is a tool of salvation, not a tool of fear.

Nighty night, sweetie, don't cheat on interviews ;)