When we perform tasks where we send data to an LLM or AI program, like generating a message to parents, we anonymize the student voice names before we send any data. When the request comes back to us, we put the names back in so that it is convenient for the teacher to use!
We ensure that each platform capability uses the best AI model to generate the strongest outputs. Socrait can be powered by various models such as OpenAI’s GPT 4o, Anthropic’s Claude, Google’s Gemini, and others (we are trying multiple models). Based on our testing, there is no one single model that is best for all tasks (yet).
Constant experimentation is being conducted to pair actions with different models to determine the most optimal AI model for specific tasks. We know that AI technology moves fast and therefore, it’s important that we’re able to flex between models to produce the best possible experience for educators. This iterative process ensures that the platform consistently delivers the highest quality responses to its users.
Comments
0 comments
Article is closed for comments.