Otter AI Chat is a powerful AI tool that provides an easy way for meeting participants to ask questions and collaborate in real-time with the transcription, all without interrupting the meeting flow. Read through the frequently asked questions (FAQs) below for AI Chat.
Disclaimers
Accuracy and reliability: While AI Chat strives to provide accurate and up-to-date information, it may occasionally generate incorrect or outdated responses. Therefore, it is important to verify any critical information obtained from this AI Chat using reliable sources.
Hallucinations: It's important to acknowledge that no vendor can guarantee no hallucinations, as they are an inherent risk with current large language models (LLM) technology stemming from the generative nature of these models. However, we have taken every precaution to mitigate this risk and ensure trustworthy outputs. Learn more about Understanding hallucinations across Otter AI Chat.
No legal or professional advice: The responses generated by AI Chat should not be considered a substitute for professional or legal advice. It is always recommended to consult with qualified professionals or experts in specific fields for personalized guidance or information.
Data privacy and security: AI Chat is designed to respect user privacy, and no personal information or data is stored or shared. However, please refrain from sharing any sensitive or confidential information during your interactions.
User responsibility: Users are responsible for their interactions with AI Chat and any actions they may take based on the provided information. The developers and owners of this AI Chat shall not be liable for any direct or indirect damages or consequences arising from the use of AI Chat.
Use at your own risk: The use of this AI Chat is at your own risk. The developers and owners of AI Chat do not guarantee the availability, reliability, or uninterrupted operation of AI Chat.
FAQs
Can other users see my AI Chat responses?
Your AI Chat conversations are private and visible only to you. Use them to explore ideas, draft content, and review insights. If you’d like to share a response, you can manually copy and paste it.
Do you use 3rd party services for AI Chat?
AI Chat uses a combination of our internal large language models and those from different service providers.
Does AI Chat send my data to 3rd parties?
Otter only sends your prompts and transcript to 3rd parties if you interact with AI Chat by sending a question. No data is sent until a user asks AI Chat a question.
Is my data stored with 3rd-party external providers?
No. Data is not stored by any of our 3rd-party external providers. The transcript and chat history are sent to external providers when a user asks a question, but the providers do not store them.
Is my data stored on Otter?
As with Otter transcripts, Otter holds your data in secure cloud storage that achieves the security standards required for our hosted platform and our customers’ needs. Otter does not access your audio recordings unless given explicit consent for troubleshooting specific product support issues and/or the user opts in to contribute data for system improvement. Read more about our data policies on our Security & Privacy page.
Is my data used for training with 3rd-party external providers?
No. Our 3rd-party AI service providers do not train on user data or use it for training.
How can I improve the AI model's responses?
Encouraging the model to respond can be effective. For instance, asking the model to “try your best” has yielded positive results for some users.
Are there any other techniques to enhance AI model interactions?
Aside from encouraging the model and managing conversation history, trying different phrasings or making your queries more specific can also yield better responses.
The answer provided has content unrelated to the transcript. How does Otter handle hallucinations?
It's important to acknowledge that no vendor can guarantee no hallucinations, as they are an inherent risk with current large language model (LLM) technology stemming from the generative nature of these models. However, we have taken every precaution to mitigate this risk and ensure trustworthy outputs.
What precautions are taken to mitigate risk and ensure trustworthy outputs?
- Carefully crafting prompts that provide clear guidance to the LLM to base its responses solely on the applicable content within Otter.
- Providing users with a feedback mechanism to report any incorrect responses, serving as feedback loops to continuously improve the models based on real-world data.
- Leveraging LLMs' reasoning capabilities to generate valid inferences that go beyond merely restating their training data.
- Ultimately, we have robust processes to ground Otter's LLM outputs in applicable content and enable continual enhancement through user feedback. Maintaining accuracy and trustworthiness remains our top priority.
Can the conversation history impact the AI’s responses?
Yes, the chat history plays a role in shaping the model’s responses. Deleting previous questions and responses and then rephrasing your question can sometimes lead to better outcomes.
Do AI Chat responses contain data or information outside of my conversations?
AI Chat answers questions based solely on the conversations in your account and those shared with you in channels. The context for answers is limited to the conversations, context (e.g., conversations, channels, and folders), and integrations you add to AI Chat.
Can I opt out of AI Chat?
Yes. Learn more about opting out of AI Chat.
Can I disable the “Otter Help” feature that provides help articles instead of directly answering my questions?
No. There is currently no way to disable the “Otter Help” feature in AI Chat. However, you can try these strategies to get more relevant responses:
- Use specific key phrases like “based on past meetings” to guide AI Chat to use the actual conversations.
- Downvote any unhelpful responses that don’t directly answer your question. This helps the AI learn which types of responses are most useful.
Feedback
0 comments
Article is closed for comments.