人工智能哲學
Issues in thePhilosophy of Advanced Artificial Intelligence
| 節 | 週五 |
|---|---|
3 10:10–11:00 | 人工智能哲學 YX206 3 節連堂 |
4 11:10–12:00 | |
N 12:20–13:10 |
* 根據陽明交大上課時間表所列
Artificial General Intelligence (AGI) broadly refers to an AI system with a range of competence comparable to human intelligence. In a hypothetical scenario in which the former becomes vastly more powerful than the latter, we speak of “machine superintelligence”. Such potential developments in AI research are increasingly an object of attention both within and outside of academia, variously eliciting excitement, fear, and skepticism. For philosophers, this represents a fertile intellectual terrain to explore. This course will examine philosophical issues pertaining to the prospect of AGI and machine superintelligence, including: 1) The notion of intelligence and its relation to language and behaviour; 2) The possibility of thinking and conscious machines, and whether thought can be reduced to computation; 3) The feasibility of AGI and superintelligence; 4) The risks presented by such developments for humanity, how seriously we should take them, and how they might be managed; and 5) The political implications of the advent of advanced AI systems.
Basic training in Philosophy English language skills
Students are asked to read one article or book chapter per week, to regularly attend class, and to participate in discussion. The learning outcomes for this course include: 1. Gain familiarity with key issues and concepts in the contemporary Philosophy of Artificial Intelligence. 2. Gain an appreciation of both the theoretical and practical implications of the advent of advanced AI. 3. Critically discuss and evaluate the main arguments and positions on the issues discussed in the course. 4. Demonstrate the ability to clearly articulate one’s own arguments and views on these issues, both in oral discussion and in writing.
Class participation (30%): Attendance (10%), oral presentations on the draft of each paper (10%), and participation in discussion according to stated expectations (10%); Mid-term paper (35%): Write a paper (min. 3,000 words long, in English) on any topic discussed between weeks 2 and 8; Term paper (35%): Write a paper (min. 3,000 words long, in English) on a different topic from the mid-term paper.
| 週次 | 主題 |
|---|---|
| 第 1 週 | Introduction: course and subject overview |
| 第 2 週 | The nature of AI and the Turing Test |
| 第 3 週 | Thinking machines: computationalism (I) |
| 第 4 週 | Thinking machines: computationalism (II) |
| 第 5 週 | Searle's critique of computationalism |
| 第 6 週 | Chalmers's defense of computationalism (I) |
| 第 7 週 | Chalmers's defense of computationalism (II) |
| 第 8 週 | Testing for consciousness in AIs |
| 第 9 週 | Discussion of mid-term paper |
| 第 10 週 | AGI & machine superintelligence: prospects & pathways |
| 第 11 週 | Skepticism about AGI & machine superintelligence |
| 第 12 週 | No class |
| 第 13 週 | AI and risk (I): should we avoid dangerous knowledge? |
| 第 14 週 | AI and risk (II): facing the control problem |
| 第 15 週 | The political philosophy of advanced AI (I): a threat of “algocracy”? |
| 第 16 週 | The political philosophy of advanced AI (II): value alignment |
| 第 17 週 | Discussion of term paper |
| 第 18 週 | Home study |
Various readings which will be shared on the E3 platform. Introductory encyclopedia entry: https://plato.stanford.edu/entries/turing-test/