I don't think that works. The fact that it can produce different output for the same input, usage of tools etc. don't really fit into the analogy or mental model.
What has worked for me is treating it like an enthusiastic intern with his foot always on the accelerator pedal. I need to steer and manage the brakes otherwise, it'll code itself off a cliff and take my software with it. The most workable thing is a pair programmer. For trivial changes and repeatedly "trying stuff out", you don't need to babysit. For larger pieces, it's good to make each change small and review what it's trying.
I feel like some of the frontier models are approaching run-of-the-mill engineer who does dumb stuff frequently. That said, with appropriate harnessing, it’s more like go-karts on a track; you can’t keep them out of the wall, but you can reset them and get them back on a path (when needed). Not every kart ends up in the wall, but all of them want to go fast, so the better defined the track is the more likely the karts will find a finish line. Certainly more likely than if you just stuck them in a field with no finish line and said “go!”.
>We would need all the intellect in the world to get the interface narrow enough to be usable, and, in view of the history of mankind, it may not be overly pessimistic to guess that to do the job well enough would require again a few thousand years.
