I was sent the link to the OpenAI Codex coding demo on YouTube, which was a lot of fun and interesting. It makes me think of the no code movement.
At one level, I really enjoyed it and I like the fluency of the API that is being used. What worries me slightly is that programming languages are being abstracted away and being put into a giant code completion tool. Although it generates new code, how new (in the sense of creative / generative) is it? If we train it on a certain set of software, are we limited to that system? As one of the speakers says, it is the world’s best pattern generator (https://youtu.be/SGUCcjHTmGY?t=933). So who controls the training and the patterns that it understands or can infer.
It worries me that it was described as proficient at coding. Arguably it may be a journeyperson/thing but it is bound by what it has learned. The other thing that niggles at me is how creative in coding might it become? What would such systems mean for creating new frameworks (not that this is always good!) or really understanding how the code works.
I get to be ‘productive’ but in a Taylorist production. My mental model becomes how to give the Codex the information to respond in a way that I want, rather than learning code or engaging with underlying technology.
There are some intriguing experiments to do with it but it also new critical approaches to these challenges.