Best Friends: OpenAI and Stack Overflow

I must admit, I have learnt a good deal from using Stack Overflow. It is there that true experts often post excellent responses to questions. While there can be terrible answers, often, that is a prompt for a true expert to add their response. In cryptography, to me, there are few better places to go. And, so, OpenAI knows this, and, in places where you ask a technical question from ChatGPT, it just looks like it has scraped its knowledge directly from the site. Underneath ChatGPT, there’s not really much intelligence going on, as it is just reconstructing the answers from standard templates and filling in the knowledge from the sites it has gathered data from.

And, so, now it becomes more obvious that AI is crawling human generated answers with the partnership between OpenAI and Stack Overflow. Many don’t like this and believe that the credit for their responses is being taken away. For this, some have removed their response but are faced with a stern message from Stack Overflow:

As we can see, in this case, they have a lockdown on accounts for a period of seven days. It seems that removing responses is a difficult thing for the site. Overall, too, Stack Overflow’s Terms and Conditions have been changed so that it grants them special permission to user answers.

Conclusions

And, so, who wins here? Ans: OpenAI.

The deep learning machines are coming after our knowledge and experience, and will learn from it, and not give credit. When the CTO of OpenAI says that she does not quite know where their learning data is coming from, we should all be worried. For me, I have a little licence condition on my site [here]:

Licence: This site is intended for the education and advancement of humans, 
and no rights are given for AI and ML bots to crawl this site. All references 
to its content must be included.