This is How Juniors Should Plan Their Tech Careers with o3 Looming🔥
Photo by David Clode on Unsplash
OpenAI’s o3 model is rather a big deal, with some calling it AGI and the biggest thing to happen to AI since GPT-3.
With this in mind, how should junior developers plan their careers? I don’t think anybody should give up the idea of working in tech, or quickly change direction to make sure they are employable. So here is what they should do.
New Day, New AI. Thinking Mode Explained
As part of their “12 days of OpenAI,” OpenAI announced its latest AI “reasoning” models, o3 and o3-mini.
AI companies are working on Simulated Reasoning (SR) models that go over and above the capability of current Large Language Models (LLM) are capable.
They’re getting close to the gold standard of Artificial General Intelligence (AGI) where a high level of human-like intelligence is achieved. One test of AGI has o3 performing better than an average human.
That’s a big deal and exciting, and is bound to have an impact on the career path of software developers.
Simulated Reasoning FTW
Other companies are developing SR models. Google announced Gemini 2.0 Flash Thinking Experimental Thursday, in November DeepSeek launched DeepSeek-R1 and the list goes on.
These models are still LLMs, but are fine-tuned to use iterative through processes. They consider their own results and use brute-force reasoning at run time.
These models are getting better with time, and soon they will perform better than most coders on most tasks.
If every car is self driving, there is still the question of where to go
So, should junior developers panic? Since it’s fun to do so, maybe.
That’s because at senior levels or above tasks tend to be complex and intertwined with the surrounding context and processes. More junior developers have more constrained tasks that future reasoning models will be able to complete with ease.
Yet there’s still hope for those junior in software development. There is no way to predict the future and know how AI will pan out. It’s good as a software developer to have both specialist and generalist skill, and if AI adapts in the future that will be something we will be able to figure out at that time.
We can also be reassured that these models aren’t close to being efficient. It costs around $3400 per task to reach a performance comparable to human, so they are simply throwing compute at the problem. This goes back to the previous point, we have time on the software development path we are currently on and if things change we can cause correct.
So, only panic if you really enjoy it.
Conclusion
If you want to become a software developer or are a junior working your way up the payscale don’t fear AI. Use it to help you improve, keep working and keep enjoying your work.
Because that’s more important than worrying about the future, right?