Why hasn’t ChatGPT been trained on data after 2021?
In: 34
chatGPT gave this answer:
1 Time of Training: The training process for large language models like mine takes a significant amount of time and computational resources. By the time my training was completed in 2021, there might not have been sufficient time or resources available to train on newer data.
2 Data Collection and Processing: Gathering and processing large datasets is a complex and time-consuming task. The training data used in my development might have been collected and preprocessed up until 2021, and updating the dataset with newer information requires substantial effort.
3 Model Update Cycles: AI models like ChatGPT go through regular update cycles, and each update typically involves a retraining process on new data. The update cycles may not have coincided with data available after 2021.
4 Maintenance and Resources: Maintaining and updating AI models like ChatGPT is an ongoing effort that requires continuous resources and attention. Prioritizing which datasets to include in updates is a complex decision-making process.
edit: for readability
Latest Answers