A few weeks ago I presented some thoughts around what I called universal ‘anything’ machines; customisable, contextually rich, personally aware characters, who are high-level experts in a wide range of disciplines, that you interact with through natural language.
- Universal ‘anything’ machines: I begin by discussing the potential of large language models (LLMs) to create and understand human-like text across different domains and media formats, and how they can be used as universal ‘anything’ machines that provide expertise on demand.
- Features and implications of LLMs: I go on to describe some of the generic and recent features of LLMs, such as multimodality, fine-tuning, embeddings, plugins, and system messages, and how they enable the creation of highly customisable, contextually rich, personally-aware characters that can interact with each other and with humans through natural language.
- Challenges and opportunities for higher education: Finally, I explore the narratives around generative AI in higher education, and how it poses both risks and opportunities for assessment, learning, and teaching. I also raise the question of the role of universities in a world where everyone has universal access to high-level expertise across all knowledge domains.
Note that the video and audio quality aren’t great. I had to do a screencast of a low-res version of the video, which then needed to be rendered again. Not much I could do about it, but thought it might be useful nonetheless.