Industry executives and experts share their predictions for 2023.  Read them in this 15th annual VMblog.com series exclusive.

Businesses will prioritize AI data quality without sacrificing time-to-market

By Sujatha Sagiraju, Chief Product Officer, Appen

In the past year, the AI industry has seen a few big challenges emerge, including issues around the quality of AI datasets and new regulations. Despite the roadblocks they faced this year, AI and data remain a top priority for businesses around the world heading into 2023. With the global AI market cap set to eclipse $300 billion by 2026, companies are looking to build successful, scalable AI deployments.

As such, we can anticipate a heavy focus within the industry on improving data quality, scaling AI with speed, and leveraging external vendors to ensure that the data used in any AI deployment is safe, scalable, and equitable. Let's examine how both issues will play out next year: 

The Battle Between Speed and Quality Will Come To a Head

For as long as businesses have leveraged AI, executives have been focused on prioritizing one of two things: the speed of AI deployment or the quality of AI data. These two have been mutually exclusive things in the past, which has led to fundamental problems in how companies build, scale, deploy, and maintain their AI systems. In the future, however, businesses will no longer find themselves in a position where they are sacrificing speed for quality or vice versa.

To avoid this problem, we will see companies continue to deploy solutions that help them both source quality data and scale AI systems more efficiently than ever before. The key to successful deployments lies at the intersection of faster deployment speed and more efficient, robust data annotation. The reality is that low-quality data will lead to low quality ML models, so data practitioners must be prepared with data sourcing, preparation and analysis.

Humans are critical to improving data quality, especially in respect to curbing bias in ML models. Technology, combined with human oversight to help spot areas of improvement along the way, will help merge speed and quality and help companies make their AI goals a reality in the coming year.

AI Data Partners Will Be Key to Accelerating AI Solutions

Given how companies are prioritizing their budgets to focus on data across the AI lifecycle, there is a huge missed opportunity when it comes to not utilizing external vendors. Data sourcing is a major bottleneck for teams building AI models. Oftentimes, organizations try to create and deploy the AI model themselves and quickly find that they lack good quantities of data, so they go to a cheaper source and end up with low-quality data. Even if an organization has access to clean, large-scale data relevant to the model, working with big data is time-consuming and requires a certain level of experience.

What these companies should have done is found an external vendor who can offer them high-quality data that enables high-performing models. Outsourcing helps to cut costs, achieve quick turnaround times, boost automation and prioritize key components like human-in-the-loop (HITL) practices. External vendors are powerful partners, and in 2023, there will be a clear shift to more and more companies looking to outsource for data preparation to help scale effectively and efficiently.

As investments in the space continue to increase and new regulations call for pivots in data sourcing and collecting practices, 2023 will be an exciting and industry-defining year for AI. The fundamentals are more important than ever, and companies will look to leverage human insight and vendor partners to drive successful AI deployments.

##

ABOUT THE AUTHOR

Sujatha Sagiraju joined Appen in September 2021 as SVP of Product and she is responsible for the product strategy. She is a technology pioneer with over 20 years of broad experience in building disruptive large-scale online services and AI/ML and data platforms. She joined Appen from Microsoft where she held leadership roles in several groups including Bing and Azure AI Platform.

Sujatha has an MBA in Technology Management from University of Washington, Seattle, MS in Petroleum Engineering from University of Texas, Austin and BS in Chemical Engineering from Indian Institute of Technology, Chennai, India.