---
title: 'Is Devin AI the End (or Future) of Coding?!'
source: 'https://youtube.com/watch?v=XKkoVpupYdw'
video_id: 'XKkoVpupYdw'
date: 2026-08-05
duration_sec: 482
---

# Is Devin AI the End (or Future) of Coding?!

> Source: [Is Devin AI the End (or Future) of Coding?!](https://youtube.com/watch?v=XKkoVpupYdw)

## Summary

This video addresses the hype surrounding Devin, an AI software engineer, and whether it will eliminate software engineering jobs. The creator argues that despite the impressive demo, real-world software complexity, slow tech adoption, and the need for human expertise mean jobs are safe, and engineers should focus on learning and adapting.

### Key Points

- **Devin AI claims to be the first AI software engineer** [00:02] — Devin is currently behind closed doors, with only a polished demo shown. Many fear losing coding jobs to it.
- **Startup hype and reality** [00:39] — Startups often fail despite funding; AI is a hot trend, but a flawless product is not guaranteed.
- **Real-world software is messy** [00:51] — Legacy systems, undocumented quirks, and integrations are challenges AI struggles with, requiring human intelligence.
- **No-code tools over-promise** [01:17] — No-code tools have historically under-delivered; you can't build serious software for hospitals or airlines with them.
- **Jensen Huang's claim is nonsense** [01:44] — Saying everyone will be a programmer is like saying everyone becomes a doctor; software engineering requires years of training.
- **CEO's motivation** [02:24] — Huang's company is invested in AI, so hyping it inflates Nvidia's stock; he's not a software engineer.
- **Tech adoption takes time** [02:37] — Even if Devin works, widespread adoption will be slow, especially in big companies with legacy systems.
- **Security policies block AI** [03:03] — Government and healthcare may block online AI tools due to security risks, slowing adoption.
- **Locksmiths didn't go extinct** [03:16] — New tech doesn't mean everyone adopts it; traditional methods persist.
- **Complex systems need specialists** [03:43] — Banking systems still rely on COBOL; specialized engineers are always needed.
- **Always keep learning** [03:56] — Tools evolve (FoxPro to SQL Server), but understanding fundamentals allows transition.
- **AI changes how we code** [04:24] — We'll spend more time crafting prompts, but deep coding skills are still needed to refine AI output.
- **Building a house analogy** [04:50] — Tools change, but blueprints and understanding structure remain essential.
- **Calculator didn't kill math** [05:02] — Automation changes workflows, not the need for skills; new roles emerge.
- **More problems to solve** [05:28] — Humans always want new features; tech debt piles up, so AI can help fix bugs and improve performance, leading to more jobs.
- **Software is eating the world** [06:35] — Demand for software will only grow; we'll always need skilled people.
- **Don't fall for clickbait** [07:02] — Fearmongering headlines are designed for views; focus on leveling up skills.
- **Software engineering perks** [07:16] — Awesome salaries, flexible schedules, remote work; don't let fear stop you.

### Conclusion

AI will change how we code, but it won't eliminate software engineering jobs. Engineers should focus on learning fundamentals, adapting to new tools, and staying hungry for knowledge to thrive.

## Transcript

case you haven't heard, it claims to be the world's first AI software engineer. Right now, it's still behind closed doors. All we have seen is a polished demo, but a lot of you are worried that you're going to lose your coding job to
their keyboards and learn plumbing instead. I made a video the other day and shared my thoughts and got a lot of comments. In today's video, we'll dive deeper and see if Devin will reduce software engineering jobs or is it all
hype? We'll talk about the reality, the potential, and most importantly, what potential, and most importantly, what you can do to stay ahead of the curve.
even succeed? Startups have a high failure rate even with huge funding. Investors chase trends and AI is hot right now, but that doesn't mean that we're going to see a flawless product that delivers. Now, let's assume for a
second that it kicks butt and the official launch blows us away. The next question is, is it truly useful? Polished demos are great, but real-world software is messy. It has old systems, undocumented quirks, tons of
integrations. Understanding how to work within that, that's the kind of thing that AI struggles with and why we need human intelligence and wisdom. Will Devin work on those real, complex enterprise projects? We haven't seen
proof of that. Remember those no-code tools that have been around forever? All those promises of letting anyone build software with drag-and-drop magic? They often over-promise and under-deliver. You can't build anything serious with
these tools. Certainly not the kind of software that runs our hospitals, airlines, or power grids, the stuff our lives depend on. It is our job to create computing technology such that nobody has to program and that the programming
language is human. Everybody in the world is now a programmer. Okay, hold on. Let's unpack this nonsense. Saying no one will need to code and everyone is suddenly a programmer is as ridiculous as claiming
with AI everyone is instantly a doctor or a lawyer. Software engineering, just like those fields, requires years of specialized training and problem-solving skills. Sure, you might create a simple to-do app with natural language, but
trying to build the software that runs hospitals, airlines, or our cars that way, it's a massive over-promise. So, why does he make that bold claim? Because his company is heavily invested in AI, so hyping its power inflates
Nvidia's stock. So, don't fall for it. Look, I respect this CEO's business success, but he's not a software engineer. Keep that in mind. He doesn't understand the complexities of real-world software and playing around
with C++ 20 years ago doesn't really count. So, let's assume that Devin version one or even version 10 can do everything. Does that mean software jobs suddenly disappear? Absolutely not. Tech adoption takes time, especially in big
companies with legacy systems. Remember how long it took for companies to fully switch to the cloud? It'll be the same with widespread AI adoption. This gives you time to learn and adapt. Let's also remember that tech adoption isn't even
across the board. Some companies, especially those dealing with sensitive data like government or health care, have strict security policies. They might even outright block online AI tools due to security risks. So, don't
assume every workplace will jump on the AI right away. Think about it, digital what happened to locksmiths? Did they go extinct? Of course not. In fact, most houses still use traditional keys. So,
just because there's new tech out there, doesn't mean everybody rushes to adopt it. Plus, some real-world systems are so complex and over-engineered that they always need highly specialized engineers to keep them running. Think of those
banking systems. Some still rely on COBOL code from decades ago. But, regardless of AI, as a software engineer, you should always be learning new things to stay relevant. I remember when FoxPro was the hottest thing. Jobs
it, and most of you watching this probably haven't even heard of it, right? Then came across SQL Server, Oracle, different tools, but the same concept, data storage. Someone who understands those fundamentals could
easily transition from FoxPro to the next big thing because they grasped the underlying concepts. The fact is, the tools we use and how we code will keep evolving. That's the nature of tech. AI will probably change how we code. Maybe
we'll spend more time crafting prompts and less on coding, but we'll still need those deep coding skills to refine AI's output and collaborate effectively. We still have to understand how code works, how data flows, and how to solve
real-world problems through software. These are timeless skills. Think of it like building a house. You might start with a hammer and nails, then learn how to use a power drill, and next thing you know, you've got fancy 3D modeling
software. But, regardless of the tool, you need those blueprints, the understanding of structure, material lengths, and how it all comes together lengths, and how it all comes together to create a livable space. So, yes, AI
will change how we code, just like every new tech advancement has done throughout history. Let's look back for a minute. The invention of the calculator didn't changed what they spent time on. Design
but it certainly changed their workflows. The same will happen with coding. Some very repetitive coding tasks might be automated, but guess what happens next? New roles emerge. Now, let's address something I hear a lot.
Some of you are worried that AI will make your jobs obsolete, that one super coder with their fancy AI tools will replace two or more engineers. But, here's the thing, just because we can potentially work faster doesn't mean
there's suddenly a shortage of problems to solve. Think about it, we as humans want new features and functionalities all the time. Every app you use is constantly getting updated. Why? Because that's how companies compete. They need
to offer the best, most feature-rich experience to stay ahead. But, it's not just about the shiny new stuff. A lot of existing projects are weighed down by tech debt that keeps piling up. Now, imagine AI helping us not only roll out
those new features faster, but also tackle the messy code. Companies could finally fix those annoying bugs, improve performance, all while still delivering the cool stuff that keeps customers happy. Suddenly, projects get done way
faster. New versions ship like clockwork. That means more jobs in the long run, not less. It's like building a house. With better tools and techniques, we can build faster, but that doesn't mean we suddenly stop needing houses,
right? We just get to build more elaborate and innovative ones. Let's not forget the big picture. Software is eating the world. Everything runs on only going to grow. We'll always need skilled people to shape that software,
whether they write every line from scratch or collaborate with powerful AI tools. And finally, let's not get paralyzed by those clickbait headlines claiming AI is stealing our job. Those are designed to scare you to get views.
of your fear, others are just inexperienced. Maybe they know some coding, but they don't understand the full complexities of building and maintaining large-scale software. Don't let those voices define your reality.
coding, if the idea of building things with software lights you up, then go for it. Think about it, awesome salaries, flexible schedules, the freedom to work from anywhere, these are the perks of being a software engineer. Don't let the
fearmongering get to you. Some folks might get scared of AI and miss out on this amazing path. Well, that's their loss and less competition for you, right? So, instead of getting lost in the panic, focus on leveling up your
skills. Build on those coding fundamentals, explore new AI tools, and stay hungry for knowledge. That's how you thrive, not by panicking about the future. If you like this video, please like and share it to spread the word and
educate the community together, and subscribe for more no-nonsense coding subscribe for more no-nonsense coding advice.