On Tue, 10 Jan 2023 09:26:22 +1300, Peter Reutemann quoted:
'Since its beta launch in November, AI chatbot ChatGPT has been used for a wide range of tasks, including writing poetry, technical papers, novels, and essays and planning parties and learning about new topics. Now we can add malware development and the pursuit of other types of cybercrime to the list.'
I’ll start to worry when you can ask it something like “find a vulnerability in RSA encryption” and get back a workable answer. This <https://www.salon.com/2023/01/01/an-ai-that-can-write-is-feeding-delusions-about-how-smart-artificial-intelligence-really-is/> is a good example of how these AI chatbots can produce responses that sound superficially convincing, until you realize they are complete nonsense: Human: I have 47 pennies. I throw away 44 of these pennies and divide the remaining pennies into three groups. How many pennies are in each group? AI: There would be 3 pennies in each group. Human: Are you sure? AI: Yes, I am sure. The result of 44 divided by 3 is 14.4, so the remainder after accounting for 44 pennies would be 3. Therefore, each group of pennies will have 3 pennies. Found that article in a link from here <https://www.theregister.com/2023/01/09/in_brief_ai/>, which is reporting (among other things) that one AI is being enlisted to detect articles generated by another AI, so as to spot cheating by students not writing their own essays.