OpenAI's Codex: Automating the Boring, Saving Programmer Time

OpenAI just released a new tool/agent/almost full software engineer called Codex, and it's insanely valuable for programmers and companies. It automated—not automates—all the boring stuff that programmers used to do. (I mean, it already did; I'm stressing this because people think it's in the future. We still have time!) Like testing each new push, making changes to typos, variable names, and functions, looking into why an error is taking place, looking at the problems faced by users in the front end and trying to find why they're arising, then writing new code for that, and then testing it. This is the kind of stuff that used to take hell a lot of time from programmers in big organizations, as there would be multiple people working on the same repo. And this tool is helping them and saving their time a lot.

The very example they gave in their demo was that this tool is part of the toolkit for every engineer at OpenAI. The question I had in my mind was, if AI gets better than engineers, why not use it themselves rather than hiring people to make it a product for the world? And this thought I had is turning into reality in the way it's supposed to: they're using it inside their company, and it's really saving them time. And one more fascinating thing is that it's still in the research preview phase of development. Just imagine or try to project how good it can become when they get feedback from actual programmers and then craft it again for the best results. And even Sam humorously tweeted, or "x'ed," what you call these days (I don't know), something like, "You want a software engineer at 20 dollars?" when people were commenting and asking why it wasn't released for Plus users.

Now I wonder and think, am I being too optimistic and trying to extrapolate it too much, or am I escaping and telling my mind not to code because AI can do this so you don't have to? Or is it a near-to-reality prediction and observation that most people fail to see as it is?

Whatever it is, one thing is true: a lot of money is being burned by humans on the plan of reaching AGI, and Sam Altman also said in an interview that they are now way more clear on how to proceed toward it.

Whatever happens, the only thing we can do is keep track of these changes and try to observe and stay updated about these things as far as possible. We should also try to project how this will touch everything and affect the market, how the market reacts, and, in the end, try to solve a problem with the help of it that people are ready to pay for, because in the end, you need to make money, and the way to do it is by solving problems.

I thought I would try it out, but it's unreachable as of now. The only thing I can do is watch YouTube and watch others using it.

Here, read the full release article from OpenAI: https://openai.com/index/introducing-codex/