Tech Job Apocalypse - AI's Chaotic Consequences
One thing is for sure, everyone expected advances in machine learning to automate away jobs but nobody was expecting artists, writers, and programmers would be the first to be impacted!
AI's impact on the job market from a software engineer's perspective
I now routinely ask ChatGPT questions about work & code. It's better than any other tool at coming up with tricky regular expressions. It can quickly format things if I give it a prompt.
I use Github Copilot when I code and it prompts me to autocomplete what it thinks I will type next. Often it is wrong but recently (within the last few months), it got better. In fact I write tests very quickly with it's help. Unless it's complex refactoring it usually helps me quite a lot.
I'll make a prediction: Any engineer not experimenting with and understanding the capabilities of ChatGPT and other LLM based AI tools is going be a 1x engineer. Every engineer that embraces these tool will be able to become a 10x engineer.
The 10x engineer thing is not a joke. It's actually more dramatic than that. The scale goes from like -10x to +100x of how productive an engineer can be. I know because I have hired a lot of engineers at this point on a variety of projects and every single person is uniquely productive across different organizations and technology stacks.
Example: In the wrong environment a good engineer can have a net negative impact. In the right environment a mediocre engineer can have a massive positive impact.
It's going to get easier to have a positive impact if you master the art of cobbling together AI helpers.
The current state of the general tech job market
Tech stocks are like a long term bond, they are interest rate sensitive and very long duration in terms of expected payoff. Welp... interest rates are up, stocks are down and the layoffs are not stopping any time soon.
Here is a secret: all the big tech companies are basically ad companies. Where do people cut spending in a recession? Marketing. Ad spend will go down. Stock based compensation will go down. Layoffs will continue until the economy improves.
Short term I only see more pain. It's incredibly hard to raise funds. Investors are pushing startups to batten down the hatches and have 2 years + of runway. If they are hiring they are doing uniquely well probably.
AI Jobs
What is an AI job? There are PhDs that research how to build new types of AI models. There is a limited supply of these because it takes 5-6 years to get a PhD.
There are a lot of software engineers who are now becoming "AI Developers". Do they know at deep levels how neural networks work? Do they know what a Transformer is? Probably not.
Does it matter?
Also probably not for what most companies want right now. They just want chatGPT plugged into their system.
Open AI, the company behind ChatGPT has an API. You can call it over HTTP with JSON. This is what backend developers do every single day, it's trivial and rote. This is how I built the recruiter email template app (which I'm selling on Acquire.com btw, I don't have much time to work on it seeing how I'm head of engineering at Bounty).
Are you looking for an AI engineer or a regular backend developer?
This is what you really need to find out now when you are talking to hiring managers.
Do they need somebody with deep understanding of building models? Or can they just get a regular old backend Java developer to plugin into the Open AI completions API and just make something previously complex start to work like magic?
I think mostly, it's the later. You are looking for a productive backend developer who is good a reading documentation to learn new APIs.
Summary
This hype is real, GPT4 is revolutionary. But what does it take to use it and embrace it? Actually not much, and when you are recruiting for "AI Jobs" just make sure to really nail down the hiring the manager on what the really want!
As always, feel free to reach out to me if you want to chat about this stuff!
- Aaron
