Name
How do we refer to the AI?
Overview:
What do we call this thing?
Some companies have started to give an actual name to their AI bot (Fin from Intercom is a good example). Even if you extend your brand to cover your assistant, we still need a common term to refer to the technology of AI itself.
Is it a bot? AI? ChatGPT? An Agent? Something else?
Users deserve to know what or who they are interacting with. In environments where a user is knowingly interacting with AI, this may not be as critical. However, consider being onboarded into a new company, and you learn you are interacting with a third-party AI agent, not someone on your team. Or, perhaps you are sharing personal information in a therapy session and you realize the "person" on the other end is AI.
This is a ambiguous pattern. We should expect to see companies play around with proprietary names, possibly keeping terms like AI associated with the bot to make it clear who the person is.
Here are a few examples in the wild today:
- Character.ai: Gives each character a name (Ask Socrates)
- Klarna: AI Assistant (Ask our AI assistant)
- Notion: AI (Ask AI)
- ChatGPT: ChatGPT (Ask ChatGPT)
- Jasper: Jasper (Ask Jasper)
- Leena: Leena.ai
- Github: Copilot (Ask copilot)
As naming conventions sort themselves out, a smaller pattern is emerging that adds a badge to the AI or the interface to making it clear when you are interacting with AI versus a human.
Intercom puts this badge inline with chats from the AI Agent. Character.ai follows the same lead. Leena adds the detail at the bottom of the screen.
This pattern is not widely adopted yet, but will likely expand to make AI interaction more transparent, to protect companies against liabilities, and to make users feel more comfortable.
Benefits:
Anti-patterns:
Uncanny Valley
Human-like names can be confusing. Avoid situations where a user would not know if they are talking to a bot or a human by adding other cues into the interface.