AI hype Is generating hundreds of billions of dollars in investment. Yet AI is still hobbled by inconsistent data, and crippled by siloed expertise. One of them-surprisingly-is that it's actually bad at simple math. Discover the other two in this insightful Forbes article.
Why is AI struggling with basic math?
AI is designed to identify patterns in large datasets rather than perform step-by-step calculations. This leads to inaccuracies in basic math tasks, which is problematic for business applications where precision is crucial. For instance, the latest version of ChatGPT achieved only 64% accuracy on math problems, which is not acceptable for critical business data.
What challenges do organizations face with their data?
Many organizations struggle with messy and inconsistent data due to years of quick fixes, acquisitions, and neglect. This 'technical debt' complicates the implementation of AI, as poor data quality can render AI systems ineffective. Experts suggest that until organizations address these data issues, the potential benefits of AI will remain unrealized.
How does siloed expertise affect AI implementation?
Siloed expertise creates barriers to collaboration between those who understand AI and those who grasp the business challenges that need addressing. This lack of communication can prevent organizations from effectively leveraging AI, as seen in cases like Novartis, where investments in technology did not translate into operational improvements until cross-functional collaboration was prioritized.
Sign in with LinkedIn