There’s been a lot of buzz about ChatGPT’s ability to respond to prompts, but can it make you a better listener?
When you start a Storied thread to listen, we use large language models, including the newest models from OpenAI, for both text summarization and content creation. Better models means a better experience. Before upgrading to the newest model, we test to make sure it would actually be an upgrade for our particular uses.
Here’s one example from a loop where a team was reporting on priorities. We noticed the new model (davinci-003) creates more detailed responses, and it also gets more things correct. davinci-002 mistakenly identifies the second part the first priority as the second priority, but davinci-003 nails both priorities correctly.
davinci-002-output: Frank is worried about two things: making incremental improvements to the standup experiment and getting all the standup threads done for the week.
davinci-003-output: Frank’s priorities are making incremental improvements to the standup experiment and learning the fundamentals of Webflow tools for user growth.
We expect these models to continue improving. For Storied’s users, this means more accurate summaries and faster content generation allowing you to both receive and share the right information, at the right time, with the right people. What becomes possible when you have 10x better clarity in your communication?
