Zeitpunkt Nutzer Delta Tröts TNR Titel Version maxTL So 11.08.2024 00:00:00 61.982 +3 3.622.897 58,5 Fosstodon 4.2.10 500 Sa 10.08.2024 00:00:02 61.979 +2 3.620.040 58,4 Fosstodon 4.2.10 500 Fr 09.08.2024 00:00:16 61.977 -1 3.617.019 58,4 Fosstodon 4.2.10 500 Do 08.08.2024 00:00:21 61.978 +1 3.614.789 58,3 Fosstodon 4.2.10 500 Mi 07.08.2024 00:00:08 61.977 +2 3.611.687 58,3 Fosstodon 4.2.10 500 Di 06.08.2024 00:00:37 61.975 +4 3.608.594 58,2 Fosstodon 4.2.10 500 Mo 05.08.2024 00:00:05 61.971 +5 3.606.116 58,2 Fosstodon 4.2.10 500 So 04.08.2024 00:00:02 61.966 +2 3.604.482 58,2 Fosstodon 4.2.10 500 Sa 03.08.2024 00:00:05 61.964 -2 3.602.552 58,1 Fosstodon 4.2.10 500 Fr 02.08.2024 00:00:23 61.966 0 3.599.813 58,1 Fosstodon 4.2.10 500
Philosopher Scholar (@philosopher) · 08/2018 · Tröts: 243 · Folger: 63
So 11.08.2024 05:15
So #OpenAI had a way to detect generated text with 99.9% accuracy for years and kept it from everyone. While there's ethics to consider that they likely withheld to gain competitive advantage, this also means model collapse can be mitigated. They CAN separate out large lengths of generated text when training future models.
https://dailynous.com/2024/08/05/open-ai-has-kept-secret-an-accurate-chatgpt-detector-for-two-years/
[Öffentlich] Antw.: 0 Wtrl.: 0 Fav.: 0 · via Tusky