HomeTechnologyComputingThe problem with AI: It's not you, it's the data

The problem with AI: It’s not you, it’s the data


Companies leave billions on the table because they can’t get their data in order. If they succeed in realizing value through data-driven initiatives such as artificial intelligencethey need to better tune and support the backend data that feeds these systems.

That’s the gist of the latest research, based on a survey of 2,500 executives and published by Infosys Knowledge Institutethat estimates that companies could collectively generate more than $460 billion in incremental profits if people managed their data resources just a little better.

This includes improving data practices, increasing confidence in advanced AI, and integrating AI more closely into business operations. Business value is still elusive.

Also: 8 ways to reduce AI burnout

The research identified three obstacles to effective AI implementations: lack of a cohesive, centralized data strategy, weak data verification, and lack of the right infrastructure. Most companies don’t have a consistent data management strategy.

Respondents want to manage data centrally, but most do not currently. Analysis of the survey results “shows that centralized data management is associated with better profit and revenue growth. 26% of respondents currently have a centralized approach; 49% would like to adopt this approach next year.

“Data is not the new oil,” emphasize the study’s authors, Chad Watt and Jeff Kavanaugh, both of the Infosys Institute. “Companies can no longer afford to think of their data as oil, which is extracted with great effort and only valuable when refined.”

Data is more like currency today: “It gains value when it circulates. Companies that import data and share their own data more widely are achieving better financial results and showing more progress in conceiving AI at enterprise scale – a critical goal for three of the four companies in the survey,” said Watt and Kavanaugh.

Also: The people building artificial intelligence are the ones who need AI the most

The success of currency depends on trust, and so does data. “Advanced AI requires trust,” the authors state. “Trust yourself and others to manage your data, and rely on AI models. Pristine data and perfectly programmed AI models mean nothing if people don’t trust and use what data and AI produce.”

Companies that have shared data, inside and outside their organization, are more likely to see higher revenues and make better use of AI, the study found. “Refreshing data closer to real time also correlates with higher profits and revenue.”

Another anti-oil analogy the study authors came up with is that data is more like nuclear power than fossil fuel. “Data is enriched with potential, needs special handling and is dangerous if you lose control. 21st century data has a long half-life. When to use it when, where to use it and how to control it are just as important as where you have to post it.”

Most companies are new to AI, the survey shows. More than 8 in 10 companies, 81%, have only deployed their first real AI system in the last four years, and 50% in the last two years. In addition, 63% of AI models operate at base capacity only and are human-driven. They often fall short on data verification, data practices, and data strategies. Only 26% of practitioners are very satisfied with their data and AI tools. Despite AI’s siren song, something is clearly missing.

Also: The true purpose of AI may no longer be intelligence

The study authors identified high-performing companies, which typically have a strong focus on three areas:

  • They transform data management into data sharing. “Companies that embrace the data-sharing economy get more value from their data,” say Watt and Kavanaugh. “Data gains value when it’s treated as currency and distributed through hub-and-spoke data management models. Companies that refresh data with low latency generate more profit, revenue, and subjective measures of value.”
  • They have made the step from data compliance to data trust. “Companies that are very satisfied with their AI (currently only 21%) have consistently reliable, ethical and responsible data practices. These terms address challenges of data verification and bias, build trust and empower practitioners to use deep learning and other advanced algorithms. to use.”
  • They involve everyone in the AI ​​process. “Extend the AI ​​team beyond data scientists. Companies that apply data science to practical requirements create value. Business leaders are just as important as data scientists. Good AI teams typically span multiple disciplines.” Data verification is the biggest challenge to moving forward, together with AI infrastructure and computing resources.
Must Read
Vulnerability tricks researchers by mimicking old threats


Please enter your comment!
Please enter your name here

Most Popular

Recent Comments