We use it to generate code in the code completion app. But sadly, with our values of trust facts and kindness, we cannot rely on large language models to comple

We use it to generate code in the code completion app. But sadly, with our values of trust facts and kindness, we cannot rely on large language models to completely generate the results as they often make up facts.