Comments on: What to (not) expect from OpenAI’s ChatGPT https://bdtechtalks.com/2022/12/05/openai-chatgpt/?utm_source=rss&utm_medium=rss&utm_campaign=openai-chatgpt Technology solving problems... and creating new ones Sun, 01 Jan 2023 14:03:44 +0000 hourly 1 By: J https://bdtechtalks.com/2022/12/05/openai-chatgpt/comment-page-1/#comment-34338 Thu, 08 Dec 2022 06:30:49 +0000 https://bdtechtalks.com/?p=15277#comment-34338 In reply to Rebel Science.

This is where humans come into the feedback loop. Even if LLMs generate data onto the internet. Making things illegal isn’t the solution to a math problem.

]]> By: Rebel Science https://bdtechtalks.com/2022/12/05/openai-chatgpt/comment-page-1/#comment-34293 Mon, 05 Dec 2022 17:54:31 +0000 https://bdtechtalks.com/?p=15277#comment-34293 Great article. In my opinion, LLMs should prohibited by law. Their potential for serious harm outweighs their utility. They will fill the internet with misinformation that looks legit. Worse, future LLMs will be trained on texts scoured from the internet, much of which will have been generated by LLMs, a dangerous feedback loop.

]]>