It’s Nearly Impossible To Get A Job in Artificial Intelligence Right Now
This is one of the most competitive fields to date.
www.forbes.com
American companies can import cheap labor but, American citizens can't import cheap electric vehicles from China lol. And the cheap labor these companies are importing seem to be messing shyt up, maybe it might be in their best interest to train the local population instead.Easier to import than train the existing population
As I said earlier could be end of year budget stuff let's see how the next two spring and summers lookMy phone been ringing off the hook since trump won. I just got a job offer and am interviewing for a side tech job. Shout out UNC
You don’t really need to “code” with AI per se. You need to know how to read documentation, and troubleshoot/research solutions to problems. I’ve used AI to build with Python, C++, Golang, Next, Svelte, Tauri, Wails and React. And I never learned any of those languages. Learning the file structure and how it all interacts is the most I’ve actually “learned” per ce. And I’m a novice. So a programmer with 10 years of experience can use AI to replace 2 - 3 junior developers working on the same project.
That’s why once I saw the industry trends last year I stopped considering looking for a job as a self taught developer. I’d be better off building my own apps and create a tech start up
Just wait until something breaksThe way folks are using AI to write code essentially guarantees that AI will indeed take over tech work. And the more users that engage in this activity, the faster this is realized.
And using AI, in the absence of real world experience, will bring you to a level that is on par with the average novice.
Not if you ask the AI to explain the code it wrote I know more about Svelte, Tauri, Golang and Wails in the two months I’ve used them, than JavaScript which I studied for over a year.The way folks are using AI to write code essentially guarantees that AI will indeed take over tech work. And the more users that engage in this activity, the faster this is realized.
And using AI, in the absence of real world experience, will bring you to a level that is on par with the average novice.
Not if you ask the AI to explain the code it wrote I know more about Svelte, Tauri, Golang and Wails in the two months I’ve used them, than JavaScript which I studied for over a year.
A lot of the times, I have to read documentation before writing a prompt because the AI doesn’t know everything. Especially if there’s a major update. A new version is Wails was released, but the AI was giving me code based on an older version. So you still have to be a good detective and show some type of technical aptitude. And it still takes hundreds of man hours.
Claude hasn't been updated since April. It was trying to tell me the version of Wails I was running didn't exist. I had to literally give it the link to the website so it could see for itself.Yeah, this is very true. Tech moves too fast. Legacy systems and knowledge are what it’s better at.
If it falls behind by a year on news/releases/etc., it falls BEHIND. They don’t call it the bleeding edge for nothing.
Claude hasn't been updated since April. It was trying to tell me the version of Wails I was running didn't exist. I had to literally give it the link to the website so it could see for itself.
I've sent it links to stack overflow, github and reddit threads, so it can find solutions to issues. Basically all the stuff normal software developers do. And all of these models get shyt wrong. And all of them are never updated on a regular basis, so you can't trust them.
Another issue i've run into is if you ask it to build a code file. It's never the same. Because there are multiple ways you can write it could literally write 200 lines of code and you ask to check it and it will keep changing it.
It's also limited in how much code it can write at one time. I think the limit is around 400 lines. So if you're writing a file that's over 1000 + lines it won't be able to write all of it. It also forgets shyt. Even if the keep the same conversation going for days, it still won't remember everything. If you ask it to revise a code file, it will some times alter the entire file. So you have to take a ton of notes during the process. When you create a new chat, you basically have to start over and explain everything in detail to get it up to speed. And then there's the tokens. You only get by so much before it kicks you out of Sonnet (Claud). Then you have to go to a lesser model and once again start from scratch.
At the end of the day, even if you're a novice using AI, it's not easy. A project that may take 2 programmers a month to complete working together, could take someone using AI 2 weeks, 1 month or 6 months to complete, depending on how they utilize their time and resources.
The first software application I built had redundancy issues. It was a patchwork mess.Let’s keep it real, AI makes it possible for an experienced programmer to complete a project in under two weeks, and in many cases, even within 48 hours.
However, just getting something built isn’t the whole picture. The real challenge is ensuring sustainability, scalability, repeatability, and reducing redundancy. AI-generated code often lacks these qualities and can come with built-in exploits or unnecessary redundancy.
I appreciate how AI is lowering barriers, but let’s be honest: a lot of bad code is being produced. It’s all fun and games until something breaks.
For context, I’m in a space where you can’t just rely on chatbot-generated code. There’s no documentation on Reddit or Stack Overflow to lean on.
Oh, and by the way, I’m also pretty sure you’re not actually developing products in the cloud.