Following @keithfrankish's tongue-in-cheek suggestion, I prompted GPT-3 to write a short philosophy paper. The result puts its strengths & weaknesses on display. The mimicry of argumentation is uncanny, if very flawed, and often unintentionally funny.

Jul 25, 2020 · 3:32 PM UTC

22
83
28
224
PDF here: drive.google.com/file/d/17yB…
1
5
10
The conclusion gets surprisingly lyrical (while calling for a machine-assisted revolution): "We will no longer be slaves to the system. We will be free."
3
1
13
If anyone has access to a plagiarism detection service like Turnitin, I'd be curious to see how much of this paper is "plagiarized" from the training data...
1
1
22
Replying to @raphaelmilliere @keithfrankish
Anyone else find this frightening?
1
Replying to @raphaelmilliere @keithfrankish
This may pass the Turing test, but it certainly won't make the Gallup mirror test...
1
Replying to @raphaelmilliere @keithfrankish
It’s like listening to a conservative relative try to make a political argument...
1
Replying to @raphaelmilliere @keithfrankish
Not sure if this proves that robots can write like humans or just that some humans write like robots
Replying to @raphaelmilliere @keithfrankish
Can you ask GPT-3 to introduce itself? "Who are you?"
1
Replying to @raphaelmilliere @keithfrankish
How was the formatting produced? Did you feed it similarly formatted text as context?
Replying to @raphaelmilliere @keithfrankish
Well now I want a Samsung Smart Toaster
6
Load more