ChatGPT Addresses the Question of Whether It Can Make a Promise
The prompt (as submitted, typos preserved):
Write an essay on whether, in genderal, an AI can fulfill the felicity conditions of performative utterances and specifically whether an AI can make a promise it is capable of keeping
ChatGPT's repsonse:
Performative utterances are a crucial aspect of human communication. They are statements that have the power to bring about a certain state of affairs simply by being uttered. The felicity conditions of performative utterances refer to the conditions that must be met in order for the utterance to be successful. One of the most important types of performative utterances is the promise. A promise is a commitment made by one person to another to do or not do something in the future. In this essay, we will explore whether an AI can fulfill the felicity conditions of performative utterances, and specifically whether an AI can make a promise it is capable of keeping.
The felicity conditions of performative utterances were first introduced by philosopher J.L. Austin in his work “How to Do Things with Words.” According to Austin, there are three main conditions that must be met for a performative utterance to be successful. First, the speaker must have the authority or ability to perform the action being promised. Second, the speaker must have the intention to perform the action. Finally, the promise must be made in the appropriate context.
An AI can certainly meet the first condition of the felicity conditions of performative utterances. Depending on the specific capabilities of the AI, it may have the ability to perform certain actions, such as sending an email or turning on a light. However, the second condition may be more difficult for an AI to fulfill. While an AI may be programmed to perform certain tasks, it does not have the same level of intentionality as a human being. In other words, an AI does not have desires, goals, or motivations in the same way that humans do. Therefore, while an AI may be capable of performing a certain action, it may not have the intention to do so.
Finally, the third condition of the felicity conditions of performative utterances may also be a challenge for an AI. Context is a crucial aspect of human communication, and it can be difficult for an AI to understand the nuances of different social situations. For example, an AI may not understand the social norms around making promises, such as the importance of following through on commitments.
In conclusion, while an AI may be able to meet the first condition of the felicity conditions of performative utterances, it may struggle to meet the second and third conditions. Specifically, an AI may not have the intention to perform a promised action, and it may not be able to understand the appropriate context for making a promise. Therefore, it is unlikely that an AI can make a promise it is capable of keeping in the same way that a human being can. However, as AI technology continues to advance, it is possible that future developments may make it more capable of fulfilling the felicity conditions of performative utterances.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.