UPDATE: In a groundbreaking statement, CEO Matt Fitzpatrick of Invisible Technologies has confirmed that humans will remain crucial in the AI training process for the next several decades. Speaking on the latest episode of the “20VC” podcast released on October 10, 2023, Fitzpatrick challenged the widespread belief that synthetic data would soon replace human involvement in AI development.
Fitzpatrick emphasized that the idea of synthetic data taking over within the next few years is a major misconception. He stated, “When I first started this job, the main pushback I always got was that synthetic data will take over, and you just will not need human feedback two to three years from now.” He argues that this perspective lacks a fundamental understanding of the complexities involved in AI training.
The CEO elaborated that the vast array of tasks requiring AI’s capabilities means that human insight will be indispensable for a long time. He pointed to industries like law, which contain extensive nonpublic information, as examples where human feedback is necessary to navigate cultural and contextual nuances. “On the GenAI side, you are going to need humans in the loop for decades to come,” Fitzpatrick declared.
Invisible Technologies, recently valued at $2 billion after raising $100 million in funding in September 2023, is in direct competition with major data labeling firms such as Scale AI and Surge AI. These companies are racing to employ millions of contractors who assist in teaching AI models essential skills, including math, science, and even emotional intelligence.
Fitzpatrick’s insights resonate with other industry leaders. Brendan Foody, CEO of Mercor, echoed this sentiment earlier in September, emphasizing that “data quality and having phenomenal people that you treat incredibly well” remain the cornerstones of successful AI training. Meanwhile, Garrett Lord, CEO of Handshake, has noted a shift in the industry towards hiring highly specialized experts instead of generalists.
The persistence of human involvement in AI training not only highlights the ongoing demand for skilled labor in tech but also underscores the emotional and economic implications for workers in this evolving landscape. As the race for high-quality data intensifies, the role of human contractors will likely continue to be a critical factor in shaping the future of AI.
Stay tuned for further updates as this story develops, and share your thoughts on the future of AI training in the comments below!
