I just refused a job at #OpenAI. The job would consist in working 40 hours a week solving python puzzles, explaining my reasoning through extensive commentary, in such a way that the machine can, by imitation, learn how to reason. ChatGPT is way less independent than people think

Jan 10, 2023 · 1:23 PM UTC

64
263
79
1,048
Replying to @DaveMonlander
How much would that pay you? And what type of commentary do they expect?
1
2
I think it would be going too far to disclose the amount. But overseas contractor-level payment. The commentary should explain the reasoning/planning before the writing of actual code.
4
1
14
more replies
Replying to @DaveMonlander
Data has to come from somewhere. Did people really think it was deus ex machina? #fauxtomation
2
3
56
What impressed me is how much information on your reasoning you need to provide. Like, write comments extensively describing what you plan to do, then write code, then run, then fail, then cross the code out, then explain what went wrong, then fix it, then run again, etc.
4
4
2
44
more replies
Replying to @DaveMonlander
Dave, can you follow me so I can DM you?
1
2
Done!
5
Replying to @DaveMonlander
Hi dave, can you please follow, me. I'd like to DM you. thanks so much!
1
4
Done
Replying to @DaveMonlander @DAIRInstitute
Cc. @sparklingruby (this might be of interest to you)
Replying to @DaveMonlander
People thought it learned this stuff without imitation? I'm confused why anyone would think this?
1
13
Load more