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

Hello there, in your opinion what would the future of humans and generative large…

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

    Hello there, in your opinion what would the future of humans and generative large language models look like? Will they compete, will they collaborate, will they just exist in different worlds?

  2. Assistant

    Here are some possibilities for the future of humans and large language models:

    • Competition: As language models continue to get larger and more capable, there is a possibility that they could compete with humans in some areas like writing, verbal communication, or other language-focused tasks. This could displace some human jobs and create economic competition. However, language models still have many limitations compared to human language capabilities, so this competition may be limited to specific, well-defined tasks.

    • Collaboration: There is an opportunity for humans and language models to collaborate. Language models could augment human capabilities by helping with writing, translation, summarization, and other tasks. Humans could help train, fine-tune, and guide language models to be more useful. This type of symbiotic collaboration could be very productive but would require careful implementation to ensure the human is still in the loop and in control.

    • Separation: It's possible that large language models and humans could continue on largely separate paths. Language models could be used in constrained, limited contexts like customer service chatbots or automated writing systems. But human language could remain primarily a human-to-human phenomenon in most areas of life. This separation would avoid competition or disruption but may limit the benefits we could gain from advanced language technologies.

    Ultimately, how this plays out will depend on how language models are implemented and applied and how well we are able to align their goals with human values and priorities. With careful management, collaboration and complementarity seem most promising but will require vigilant oversight to avoid potential issues like competition over jobs or unintended consequences from advanced language systems. But there are still many open questions about how far language models may advance and what their limitations will be relative to human language.

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