Major AI firms, including Scale AI and Turing, are transitioning from low-cost data labeling workers to more skilled professionals in response to the growing demand for enhanced model quality.
Changes in Data Processing Approaches
Previously, data labeling workers were tasked with simple annotation jobs. However, as demands for data volume have increased, companies recognized the need for more qualified support. This has shifted the focus towards specialists in fields such as biology and finance.
New Opportunities for Specialists
Leading companies like Scale AI, Turing, and Toloka have begun hiring experts with specialized knowledge. Olga Megorskaya, CEO of Toloka, noted that 'the AI industry has long overlooked the importance of data for model training,' resulting in increased investor interest in these firms.
Financial Aspects and Market Impact
Major firms like Turing are compensating their experts 20-30% above current pay rates. While AI companies allocate only about 10-15% of their budgets to data, this still translates to substantial sums in the context of their spending on computational resources. Jonathan Siddharth, co-founder of Turing AI, emphasized the necessity of using data from real human experiences to improve model quality.
The shift towards employing more qualified data labeling specialists highlights the growing importance of this task within the AI industry. This may lead to improvements in model quality and their ability to learn from more accurate information.
