The Role Of Synthetic Data In Expanding AI

A huge limitation to the development of effective AI at the minute is access to relevant and risk-free data. This is where synthetic data can help, Analyst Rowan Curran discusses how synthetic data can help accelerate AI efforts.

Rowan starts with a definition of what synthetic data for AI means. He points out that synthetic data generated for AI is different than synthetic data used for load-testing or performance-testing. Rowan goes on to say, “We’re talking about data sets that mimic real world data” He continues to say, “ There is just not enough data of the right type or quality to infer and predict the things we want to predict.” Curran emphasises that the synthetic data is not “fake” but is “synthesized” for a specific use”.

There are some key advantages to using synthetic data to test AI models over simply encrypting or anonymising actual data, as synthetic data doesn’t actually represent a real person’s identity or trait, there’s no risk of releasing personal data accidently or through an attack. Although inference attacks aren’t common, they can be used to infer certain things about real data that sits behind an AI model, this is especially valuable in the healthcare sector where patients’ data is used. Rowan also mentions that synthetic data can help alleviate governance concerns around sharing personal data (such as patient or customer data) between business partners.

Following on from this, the discussion develops into how synthetic data for AI is created. In some cases, it’s an assumption of an existing data set to create a much bigger one that closely mirrors the original but is not actual personal data. There are also platforms that will generate synthetic data based on a specific parameters or inputs, this is useful in the computer vision applications where a user may want to generate a 3D object for a game or virtual world.

During this discussion, Curran provides plenty of specific examples of practical uses for synthetic data and even some cases where it might not be the best solution.

For the full article, please click below:

The Role Of Synthetic Data In Expanding AI – Forrester

For the latest industry news, click here: