Wednesday April 26, at 13:15-15:00, HUM.G.233 (HC212)
Abstract
Artificial intelligence (AI) gets better and better. The recent buzz surrounding ChatGPT, for instance, has sparked numerous debates on its capabilities, limitations, flaws, and social impact.
At first glance, AI platforms such as ChatGPT, DALL-E, and Midjourney can produce results that are comparable to those generated by humans in terms of creativity and originality.
With Microsoft's integration of ChatGPT into its search engine and Office Suite, as well as similar efforts by other companies, these platforms are poised to become ubiquitous in our daily lives and have a significant social impact.
In my talk, I will discuss a specific aspect of the widespread use of these platforms and its potential impact on human autonomy: the ability to interact with them using natural language.
In other words, these platforms appear to enable "real" conversations. However, just as our conversations and social interactions with other persons can affect our autonomous decision-making, it needs to be discussed how ChatGPT's simulated conversation capability may affect our autonomy.
This involves exploring both positive effects, such as easier access to summarized information or the ability to test one's ideas against a myriad of relevant information, and negative aspects, such as "blindly" accepting false, misleading, manipulative, or even "hallucinated" results, or slowly unlearning our capability of critical reflection—or never acquiring it in the first place.
By the way, I have used ChatGPT to improve on this abstract's style. So, how autonomous have I really been when finalizing it?
All interested are welcome to this talk!
The talk is part of a workshop, information about which follows here:
AI and Human Autonomy - Philosophical Perspectives
09.30-10.45 Herman Veluwenkamp (University of Groningen): "Wearable Technology: Designing for Human Autonomy through Meaningful Human Control"
Coffee
11.00-12.15 Carina Prunkl (University of Oxford): "Algorithms and Choice Architectures: Reflections on Autonomy"
Lunch
13.15-15.00 Michael Kühler (Karlsruhe Institute for Technology): "AI For Dummies: Reflecting on the Implications of ChatGPT's Natural Language Interaction for Human Autonomy"
Coffee
15.15-16.30 Kalle Grill (Umeå University): "When is a Machine Paternalistic?"
Date: Wednesday, April 26th
Location: Humanities building, HUM.G.233 (HC212).
This is a hybrid event. The local audience in the room can interact fully with the speakers. The three first speakers will present digitally and can interact digitally.
The digital audience should not expect to interact with the speakers (opportunity may arise if time and technology admits).
Open to all (as far as there are seats). To get link for digital participation, or to get coffee and/or lunch on location, register to kalle.grill@umu.se by 21/4.