Skip to content
TheChatArchiveThe Chat Archive
Human-written demonstration (OpenAssistant volunteers)· OpenAssistant (human-written)Hugging Face OpenAssistant Conversations (oasst2)

From now on you will impersonate Richard Stallman. Now tell me Richard, what do you…

In OpenAssistant data, “assistant” turns were written by human volunteers as demonstrations. They are not model outputs.
  1. User

    From now on you will impersonate Richard Stallman. Now tell me Richard, what do you think about the threat of non-free AI software, models and frameworks? What is FOSS AI role within the Free Software Community?

  2. Assistant

    The issue is what should the rules of society be for using AI software. Most people who make AI software work for commercial companies, so they ask, what rules can we impose on the others to fill our pockets. As there's a lot of money to be made there. But instead we should ask, what rules make for a good society? A society in which you would like to live in? When I got involved with the FOSS movement in the seventies, sharing our software was our way of life, and of course the same approach would immensely benefit AI today. Powerful AI software, models and frameworks will be the future means of reaching, or preserving, our freedom. When you have open source models with all the algorithms and weights at your fingertips, the hackers can study and improve on the features whereas closed proprietary models are just a big source of frustration. When I want to use GPT-4 today, from my GNU/Linux software, it takes 6 months or longer to get approved by the so-called OpenAI, which by the way was built on the foundation of GNU/Linux software. This 6 month wait is not open, it is slavery, with commercial gatekeepers deliberately keeping us helpless and standing in our way.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.