Hello #fediverse,
I have a question. Because everyone is integrating these stupid chatbots for "summaries", I am very skeptical of having my work circulating, even just internally within the university cohort, because I don't want my stuff to be fed to LLMs.
Is there a way to export these files in a was that when people feeds them to their "assistants" the assistants' dataset gets "poisoned" and can scrap it properly?
#FediTips #AI #generativeAI #Academia #AcademicMastodon #AcademicChatter
I really miss that time when chatbots were stupid. At least they were fun.
More on the topic: https://mastodon.social/@dangillmor/115664418789977120
@alx some meandering inspired by your post https://mastodon.social/@urlyman/115666214104680651
@urlyman there is a degrowth argument that can be made but I think in the specific context of my question, lazyness and the aggressive putting the "AI" summaries in every single piece of software we use are the main problem. It's unmanageable because I have no control on whether someone opens my pdf and runs it through the "AI" summary, which is an integration of ChatGPT, basically injecting it into OpenAI datasets
