Tag: ChatGPT
-
Language models don’t sometimes hallucinate. They always hallucinate.
By now, most people have come across the issue of language models like GPT hallucinating, where the model generates an output that’s unrelated to the prompt. Or, you may find that the generated responses increasingly diverge from the topic (as the error rate in the model accumulates over increasingly long sessions). When the response generated…
-
Universal ‘anything’ machines – North-West University symposium
In this presentation I describe the concept of “universal ‘anything’ machines,” which leverage large language models (LLMs) capable of producing human-like text across various domains. These LLMs enable the creation of customisable, context-aware characters with expertise in multiple disciplines, accessible via natural language. The talk also addresses the implications for higher education, discussing the potential…
-
Example of a tutorial session with ChatGPT as the tutor
I wanted to do a little experiment using ChatGPT as a tutor, after coming across a project for building a personal AI tutor (see my last weekly digest). I’ve posted the transcript of the exchange below. A few points of interest I thought worth noting: When reading through the exchange below, remember that ChatGPT only…
-
How I used to get an overview of a new topic, compared to using ChatGPT
The other day I was in a meeting where participants were talking about different routes our institution offers towards a PhD, and I realised that I didn’t have a strong sense of the differences, and relative merits of each route. What I would usually do in a situation like this is something like the following:…
-
Internal calibration through ChatGPT
in AII’ve been using ChatGPT for internal calibration of my position with respect to certain topics where I know a wide range of perspectives exists. When I want to know how far from centre I am, I query ChatGPT to find out what the ‘average’ position is. Since ChatGPT can only be the average of what…
-
Rejected AMEE abstract (workshop) | ChatGPT and its implications for health professions education
See this brief post on my reasons for sharing rejections. Background The release of ChatGPT in November 2022 was accompanied by an explosion of interest in both mainstream and academic media. It has simultaneously been described as the saviour of higher education, and the beginning of the end. The fact is, it could be both.…
-
Podcast | Lex Fridman interviews Sam Altman (OpenAI)
I really enjoyed this conversation between Lex Fridman and Sam Altman. Sam is the CEO of OpenAI, who developed the GPT large language model, and the ChatGPT interface. I thought that his responses were thoughtful and considered. The conversation veers toward the technical, but there’s enough at a high level for this to be useful…
-
Conversation with Dave Nicholls on generative AI
A few weeks ago I had a conversation with Dave Nicholls for his Paradoxa substack, on the topic of large language models and other forms of generative AI. Dave and I wanted to talk about the implications of generative AI, and of ChatGPT in particular, on higher education and professional practice. In addition to our…
-
Generative AI in higher education – HELTASA conference
This is a recording of a presentation I gave at the HELTASA conference, on the topic of generative AI in higher education. The focus of the talk was on ChatGPT but I tried to think about generative AI more broadly, including image generators like Midjourney and Stable Diffusion. I also tried to avoid making this…