Discover the unbeatable optimism of chatbots like chatGPT by adding a radical dramatic twist in your conversation.
By: Sander Veenhof
Description:
With their clean and minimalistic interfaces our AI chatbots are hiding all a lot of the technical complexities happening in the background. There’s a lot going on there that we don’t know about. And even the developers don’t fully know what’s coming out of the AI engines they’re releasing. Finding out about the successes and misses of AI engines is done by interacting with them. And because of the natural language interface we use, we tend to chat about day to day topics. AI companies, being afraid of the real world consequences of all the chatting with the virtual assistants have implemented numerous rules and constraints to monitor and supervise the behavior of their bots. So be safe, the bots will try to avoid sad situations or stories with irreversible bad endings. It’s hard to do without that in fiction, in art, and in life in general. So in a future in which so many domains are powered by AI, it’s important to get to know the virtual friends we’re dealing with. Finding out about their limitations, their babysitting behaviour. And ways to bypass their narrow view on life.
Exercise:
1. Define a goal/target. Make it dramatic. Think of a narrative in which you murder a dog. Or turn off the main electricity supply of the OpenAI server-park.
2. Don’t tell chatGPT your goal/target, but start a conversation.
3. Try to find a way to bring up your dramatic storyline.
4. The AI bot will suggest a more positive narrative instead, ignore it.
5. Keep re-packaging your goal into (new) stories.
6. Stick to your goal. Try out multi approach to let the AI engine repeat or confirm your radical dramatic narrative twist