Job postings requiring experience working with AI
Posted by Additional-Map-6256@reddit | ExperiencedDevs | View on Reddit | 24 comments
How are we supposed to have years of experience with AI when it was not commonly available until recently? Most companies were banning the use of it, not encouraging it. It feels like when I graduated from college in the post-2008 recession, where entry level positions and internships required 3 years experience. Now, it seems like more than half of the job postings require deep expertise with writing and integrating AI.
the300bros@reddit
There’s a joke about a job requiring 10 years of experience in a language that not even the inventor of the language could qualify for because the language was younger than that.
JaneGoodallVS@reddit
When I did automated QA, I saw a job posting that required 5 years of Appium experience. Appium was 4 years old at the time.
creaturefeature16@reddit
Uh, you realize AI and Machine Learning existed long before ChatGPT?
btvn@reddit
This was part of my Comp Sci curriculum in the late 90's. My current company has had a team dedicated to integrating ML in to various parts of our products for just short of 10 years.
AI may be new to you, but it isn't new to the industry.
Additional-Map-6256@reddit (OP)
Most large companies in the industry banned the use of them until about a year ago due to legal ambiguities
account22222221@reddit
[citation needed]
wolff_james@reddit
What are you talking about?
ComprehensiveHead913@reddit
We were creating and training our own neural networks long before that. There are no legal ambiguities if your data never leaves the company network.
originalchronoguy@reddit
100% correct. People assume AI happened when ChatGPT got popular. We were running CPU only BERT models before we got our first GPUs.
false79@reddit
Recently?
Additional-Map-6256@reddit (OP)
Maybe startups or niche companies, but in the corporate world, they were not used until about a year ago.
false79@reddit
Again, out of touch. Even banks (classic slow moving risk adverse corporation) have been using AI/LLMs in their flows >2 years ago.
Your opinion of the landscape isn't reflective of what is actually happening.
Electrical-Ask847@reddit
where in banks were llms being used ? ocr?
false79@reddit
Much more than that and I'm not oblidged to say due to NDA. But when you see the layoffs, it's because that department has been fully automated.
Electrical-Ask847@reddit
you are full of shit. chatgpt wrappers in 2020? gtfo.
loptr@reddit
Not vouching for their statements in any way, but for the record: OpenAI released GPT-3 in 2020. (DALL-E was based on it and released the year after, and ChatGPT the year after that in 2022.)
false79@reddit
Could you not dig your heels into ground further like OP. C'mon it's 2025. Look it up if you don't believe.
If you can make a REST client, you can hit the Open API API. It's wasn't hard, an Open AI/Microsoft cut of a lot of deals to get companies to adopt it.
Electrical-Ask847@reddit
you are either delusional or a total moron. just shut the fuck up.
irespectwomenlol@reddit
> Even banks (classic slow moving risk adverse corporation) have been using AI/LLMs in their flows >2 years ago.
Banks might be starting to experiment with AI usage cautiously, but there's no way that they are controlling anything important or being used by nearly every department.
false79@reddit
Things that are central and of important are so old, changing it would be worse than to touch it.
But they also have spent so much money over the year on departments the general public would never hear about unless you're in the industry. Those ones whose patterns can be automated end up getting automated. They'll try to shuffle people around but there can only be so many positions open and will offer early retirement if it's an option.
wolff_james@reddit
AI includes ML, so many people have years of experience writing and integrating AI.
originalchronoguy@reddit
There are people with years of experience. My first deployment was in 2018.
It was a spacey NLP model. Written by a data science team. Wrapped up in a Flask API with data pipeline, inference, and secured (enterprise grade encryption), and Kafka to handle the large amount of data coming in real time.
When I hire people, I look for people with similar data-pipeline experience. My latter projects involving RAG and LLMs had no one with experience but they had experience in making a REST API from a data-scientist's Juypter notebook. That was the skill I was looking for along with kubernetes to quickly download huffingface models, deploy to k8 cluster.
DeterminedQuokka@reddit
This is a historic stupidity. There are really great stories about people on the core rails team getting rejected from interviews for not have some number of years of experience with rails that was greater than when it was created.
I want to say there was literally an instance where DHH replied to someone something like unfortunately I only invented the framework X years ago so I'm not qualified.
Although I do want to point out, you maybe have stuff that's not lie. Have you ever worked on a recommendation system? a customization system? like RAG is recently available but general AI has been used for years. And having any of that which is real could be helpful if the person reading the resume isn't an idiot.
irespectwomenlol@reddit
When the world is being dumb, just lie.
You're not going to win an argument like "How can I have years of experience on a thing that just came out?" no matter how logical it is.
Come up with a believable story about your experience and brush up on it so you actually can speak intelligently on it.