THE DRAWBACKS OF CHATGPT

I’ve been thinking about OpenAI’s novel “conversational agent” so-called ChatGPT’s pros and cons. I am sure most of you have been mesmerized by the fluency and efficiency of ChatGPT but, as an IP lawyer, I am drawn to discover what might be the drawbacks of this innovation.

So far, as ChatGPT are being trained on all sorts of text from the internet, such as Wikipedia, blog posts, books, and academic articles, it is highly probable that it may eventually and indispensably violate copyrights of authors and artists whose works are being exploited by ChatGPT without having their prior consents.

This is due to the fact that ChatGPT needs collective writing of humans across the world, past and present which may be protected under copyright laws. One can argue that ChatGPT comes up with novel ideas and writings depending on the existing literature on the internet which is, by the way, leading to a fair use but this is only one side of the story. The other side of the story is that ChatGPT must, in any case, be trained to respect IP rights of others.

You may read about other drawbacks of ChatGPT via https://lnkd.in/ds_c4wEs and https://lnkd.in/dYNUw5tG