Recently, Appen's 2024 State of Artificial Intelligence Report has garnered significant attention. Based on a survey of over 500 U.S. IT decision-makers, the report reveals a 17% growth in generative AI over the past year. However, this surge in growth has also brought considerable challenges in data management for businesses.
Si Chen, Strategic Head at Appen, mentioned in an interview that as AI models are increasingly applied to more complex and specialized fields, the demand for data is rising accordingly. Merely having data is insufficient; businesses require data that is not only accurate and diverse but also clearly labeled and suitable for specific AI usage scenarios.
The application of generative AI is expanding, from IT operations to research and development, with companies leveraging this technology to enhance efficiency. However, despite the thriving development in this field, the return on investment for AI projects is declining.
Since 2021, the proportion of successfully deployed AI projects has decreased by 8.1%, and the number of projects yielding substantial returns has also dropped by 9.4%. This is primarily due to the increasing complexity of AI projects, as many businesses are now venturing into more challenging applications of generative AI, which requires higher quality data support.
Additionally, the report points out that data quality issues are becoming more severe. Since 2021, the accuracy of data has decreased by nearly 9%. Currently, 86% of enterprises need to update their models quarterly, which demands higher accuracy and diversity in data. To address these issues, many businesses are turning to external data providers for assistance.
Moreover, data preparation has become the biggest bottleneck for enterprise AI projects. As the complexity of generative AI models increases, businesses need a more strategic approach to ensure data quality and consistency. Furthermore, human intervention in this process has become increasingly important. Surveys show that 80% of respondents believe that "human-machine collaboration" in machine learning is crucial. This model not only helps improve the performance of AI models but also ensures their ethicality and relevance.
Key Points:
🌟 Generative AI has seen a 17% growth in 2024, yet businesses face significant challenges in data management.
📉 The proportion of successfully deployed AI projects is declining, along with the return on investment.
🔍 Data quality issues are becoming more severe, necessitating higher quality data to meet the demands of complex AI models.
