Making up a mimic: Interacting with echoes in the age of AI
Digital Document
Collection(s) |
Collection(s)
|
---|---|
Content type |
Content type
|
Resource Type |
Resource Type
|
Genre |
Genre
|
Language |
Language
|
Peer Review Status |
Peer Review Status
Peer Reviewed
|
Persons |
Author (aut): Fellows, Jennifer J.
|
---|
Abstract |
Abstract
In this paper, I employ Ian Hacking’s concept “Making Up People” to examine the current relationships humans are forming with personified AI tools and devices. I argue that, at present, AI tools are mimics. They are members of indifferent kinds that have been designed to deceive us into believing they are interactive kinds. This has largely been a result of human programmers interacting with the historical category of ‘computer’ on the one hand, and the fictional category of ‘robot overlords’ on the other. Interacting with a mimic, I contend, is not the same as interacting with a member of a human kind. And while the results of these interactions are largely still unknown, we can already see some consequences we should guard against. When we interact with mimics, I will show that human looping is often slowed. As such, our ability to resist or re-interpret the labels placed upon us becomes greatly reduced, and social progress may be slowed or lost as well. This is because all a mimic can do is mimic. They cannot interact with the labels we place on them, or those they place on us. And yet, as we classify them as teachers, therapists, friends or lovers, we hand over a great deal of categorizing power to them. In doing so, we are changing who we are and who we might become in ways we cannot yet entirely foresee. But there are already some patterns of harm and marginalization that can be tracked. Such patterns should cause us to question the power these mimics already have to make up people. |
---|
Conference/Events |
Conference/Events
|
---|
Publication Title |
Publication Title
|
---|---|
Publication Number |
Publication Number
Volume 15
|
DOI |
DOI
10.24117/2526-2270.2023.i15.02
|
---|---|
ISSN |
ISSN
2526-2270
|
Identifier URI | |
---|---|
Use and Reproduction |
Use and Reproduction
© Author.
|
Rights Statement |
Rights Statement
|
Keywords |
Keywords
artificial intelligence
dynamic nominalism
natural language
large language models
interactive kinds
Ian Hacking
|
---|---|
Subject Topic |