From the course: Implementing a Data Strategy for Responsible AI
Join today to access over 25,600 courses taught by industry experts.
Scalability and efficiency changes
From the course: Implementing a Data Strategy for Responsible AI
Scalability and efficiency changes
- [Narrator] Regardless of the AI product type you're building, you're regularly weighing the scalability possibilities with their corresponding efficiency impacts. Hopefully, you're not making these decisions alone. Input from your data and AI practitioner coworkers would make it easier to decide and set up more consistent practices. Balancing scalability and efficiency is rooted in avoiding garbage in garbage out. Commonly garbage in, garbage out refers to input data being of poor quality. But for generative AI, it also includes effective data management and observing computational limits. You have to ensure that you don't overload any AI model with data, especially during the test and validation phase. The last thing you want is for the ingestion of too much data to stall your data processing progress, or to introduce data incompleteness given the dropping of data packets so you slowly increase data workloads while checking that there's little to no performance degradation. In…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.