From the archive: What does ChatGPT really know? - Many Minds
- Rating
-
- Fav.
- Waveform
-
- Action
-
- Link
- Title
- From the archive: What does ChatGPT really know?
- Description
- <p>Hi friends, we're on a brief summer break at the moment. We'll have a new episode for you in August. In the meanwhile, enjoy this pick from our archives!</p> <p>----</p> <p><em>[originally aired January 25, 2023]</em></p> <p>By now you’ve probably heard about the new chatbot called <a href="https://openai.com/blog/chatgpt/">ChatGPT</a>. There’s no question it’s something of a marvel. It distills complex information into clear prose; it offers instructions and suggestions; it reasons its way through problems. With the right prompting, it can even mimic famous writers. And it does all this with an air of cool competence, of intelligence. But, if you're like me, you’ve probably also been wondering: What’s really going on here? What are ChatGPT—and other large language models like it—actually doing? How much of their apparent competence is just smoke and mirrors? In what sense, if any, do they have human-like capacities?</p> <p>My guest today is <a href= "https://www.doc.ic.ac.uk/~mpsha/">Dr. Murray Shanahan</a>. Murray is Professor of Cognitive Robotics at Imperial College London and Senior Research Scientist at DeepMind. He's the author of numerous articles and several books at the lively intersections of artificial intelligence, neuroscience, and philosophy. Very recently, Murray put out a paper titled <a href= "https://arxiv.org/abs/2212.03551">'Talking about Large Language Models</a>’, and it’s the focus of our conversation today. In the paper, Murray argues that—tempting as may be—it's not appropriate to talk about large language models in anthropomorphic terms. Not yet, anyway.</p> <p>Here, we chat about the rapid rise of large language models and the basics of how they work. We discuss how a model that—at its base—simply does “next-word prediction" can be engineered into a savvy chatbot like ChatGPT. We talk about why ChatGPT lacks genuine “knowledge” and “understanding”—at least as we currently use those terms. And we discuss what it might take for these models to eventually possess richer, more human-like capacities. Along the way, we touch on: emergence, prompt engineering, embodiment and grounding, image generation models, Wittgenstein, the intentional stance, soft robots, and "exotic mind-like entities."</p> <p>Before we get to it, just a friendly reminder: applications are now open for the Diverse Intelligences Summer Institute (or DISI). DISI will be held this June/July in St Andrews Scotland—the program consists of three weeks of intense interdisciplinary engagement with exactly the kinds of ideas and questions we like to wrestle with here on this show. If you're intrigued—and I hope you are!—check out <a href= "http://disi.org/">disi.org</a> for more info.</p> <p>Alright friends, on to my decidedly human chat, with Dr. Murray Shanahan. Enjoy!</p> <p> </p> <p>The paper we discuss is <a href= "https://arxiv.org/abs/2212.03551">here</a>. A transcript of this episode is <a href= "https://drive.google.com/file/d/1HQIV90Ej0TkQQRb3DdsQMhKbTZPiLKuQ/view?usp=share_link">here</a>.</p> <p> </p> <p><em>Notes and links</em></p> <p>6:30 – The <a href= "https://proceedings.neurips.cc/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf">2017 “breakthrough” article</a> by Vaswani and colleagues.</p> <p>8:00 – A <a href= "https://www.nytimes.com/2020/11/24/science/artificial-intelligence-ai-gpt3.html">popular article</a> about GPT-3.</p> <p>10:00 – A <a href= "https://www.nytimes.com/2022/12/16/opinion/conversation-with-chatgpt.html">popular article</a> about some of the impressive—and not so impressive—behaviors of ChatGPT. For more discussion of ChatGPT and other large language models, see another <a href= "https://www.youtube.com/watch?v=BqkWpP3uMMU&t=2580s&ab_channel=MachineLearningStreetTalk">interview</a> with Dr. Shanahan, as well as interviews with <a href= "https://www.youtube.com/watch?v=VDSw7np2ovM&ab_channel=TheTWIMLAIPodcastwithSamCharrington">Emily Bender and Margaret Mitchell</a>, <a href= "https://podcasts.apple.com/us/podc
- Publication Date
- 2024-07-24T11:00:00+00:00
- Status
- completed
- Website
- https://manyminds.libsyn.com/from-the-archive-what-does-chatgpt-really-know
- Length
- 55:09
- File
- /podcasts/Many Minds/1721818800-4851.mp3
- Size
- 75.75 MB
- Bitrate
- 187-CBR
- Channels
- 1
Queries: 25 | Cache Hits: 0 | Load Time: 0.0526 | 2 MB