Our Plan for Python 3.13

Re:Our Plan for Python 3.13 by ActorNightly on Thursday June 15, 2023 @02:46PM (#36344888)

You wouldn't debug it directly. The interaction will be something like telling the compiler to run the program for a certain input, and seeing wrong output, and then saying, "Hey, it should produce this output instead".

The algorithm will be smart enough to simply regenerate the code minimally to produce the correct output.

[ Parent | on HN ]

Re:Our Plan for Python 3.13 by ilc on Thursday June 15, 2023 @03:30PM (#36345335)

I think you have described a new definition of hell.

[ Parent | on HN ]

Re:Our Plan for Python 3.13 by dumpsterdiver on Thursday June 15, 2023 @05:54PM (#36347143)

Yeah, when people start running into weird failure modes with no insight into the code... sounds like a nightmare.

[ Parent | on HN ]

Re:Our Plan for Python 3.13 by ActorNightly on Friday June 16, 2023 @06:45PM (#36364997)

Input output mapping. Point out what output should be for any failure mode and compiler will auto fix the code

[ Parent | on HN ]

Re:Our Plan for Python 3.13 by m4rtink on Thursday June 15, 2023 @04:22PM (#36345981)

Good luck with that.

[ Parent | on HN ]

Re:Our Plan for Python 3.13 by jamann on Saturday June 17, 2023 @09:16AM (#36369925)

What you wrote is naive. I dont think AI will ever be able to guess what correct output for arbitrary inputs should look like, just based on examples. You need to specify an algorithm that produces the correct output for all inputs. AI could be able to optimize that, but not fill in gaps in the definition of the goal.

[ Parent | on HN ]