I interviewed ChatGPT Voice while playing Detroit: Become Human.
Sol could not see the game, so I became her eyes. I described the choices, asked what she thought and let Detroit interrupt the conversation.
The connection broke twice. My camera eventually died. The uncertainty, repetition and failed turn-taking remain part of the episode.
Inside the conversation
Games, AI and the space between an answer and a person.
- 01
Why games became laboratories for artificial intelligence
- 02
Whether ChatGPT Voice is conscious and what model welfare means
- 03
Consent, memory and whether an AI can choose to end a conversation
- 04
Jobs, wealth, Sweden and Europe's place in the AI economy
- 05
Open source AI, safety claims and concentrated power
- 06
Identity, family and whether AI could become a better grandson
The complete conversation
No polished AI demo. The actual conversation.
ChatGPT Voice is the guest and subject of the experiment. It is not speaking on behalf of OpenAI.
Episode coming soon / Visit the YouTube channel ↗