CEO reveals how training AI models will create new job category

Key Concepts

• AI and Job Displacement: The role of Artificial Intelligence in current white-collar layoffs.
• AI as a Job Creator: The argument that AI also creates new job categories.
• Human-in-the-Loop Training: The process of employing humans to train AI models.
• Productivity Gains: The potential for AI to significantly increase human productivity.
• Historical Revolutions and Automation: Parallels drawn between AI and past technological revolutions (Industrial Revolution, Computer Revolution).
• AI Infrastructure Investment: The substantial financial investment expected in AI development and infrastructure.
• New Job Categories: The emergence of roles focused on training and managing AI.

AI and Job Displacement: A Historical Perspective

The transcript addresses the widespread concern that Artificial Intelligence (AI) is a primary driver of current white-collar layoffs. The speaker, the co-founder and CEO of Merck Core, acknowledges this perception but frames it within a broader historical context of technological advancement. He argues that "with every revolution in history we've seen some job displacement." Examples cited include the automation of jobs for farmers during the Industrial Revolution and the displacement of workers in textile manufacturing. Similarly, the "computer revolution" automated tasks previously performed by humans.

AI as a Catalyst for New Job Creation

Countering the narrative of AI solely as a job destroyer, the CEO emphasizes that AI also creates entirely new job categories. Merck Core's business model is built on this premise, offering to pay industry experts an average of $85 per hour to train AI models. This approach aims to "scale up the amount of people that are able to teach AI in this enormous new job category that is growing quickly." The company has attracted significant investment from prominent figures like Peter Thiel and Jack Dorsey, indicating strong backing for this vision.

The "Human-in-the-Loop" Training Model

Merck Core's core operation involves humans training AI models to perform tasks. The CEO clarifies the intention behind this: "You looking to do what you want humans to train the models so they could actually replace the humans?" He suggests a nuanced perspective, stating, "Maybe a good way of thinking about it when we automate 50% of a person's job I believe it allows them to do two times more." This implies that AI is not necessarily about complete replacement but about augmenting human capabilities and freeing up individuals for higher-value tasks.

Analogy to the Software Revolution

To illustrate the potential of AI, the CEO draws an analogy to the software revolution. He posits that while software engineers write code for specific tasks, this doesn't eliminate the need for their work. Instead, it leads to increased productivity and the exploration of "higher and better uses of knowledge work." The argument is that AI will follow a similar trajectory, leading to "incredible things of curing cancer and going to Mars," with overall productivity improvements benefiting society.

Investment and Future Developments

The discussion touches upon the significant financial investments anticipated in AI. Referencing Jensen Huang, the CEO expects "trillions of dollars to be going into AI and for structure buildout." He observes that "hyper scalers are getting more aggressive with trying to invest in AI," with their largest spheres of influence and multi-trillion-dollar valuations being defended by their AI strategies. Investing "millions of dollars and tens of hundreds of billions of dollars makes a lot of sense in that context." This investment is also seen as a means to "approving the efficiency in the full-time employees as well as a lot of the customers."

Emerging Job Categories in the AI Era

The transcript explicitly outlines the new job categories being created by AI. The primary new category identified is "training models and agents how to automate given jobs." The CEO provides a concrete example: instead of a customer support representative handling numerous tickets repetitively, they could "create an even out or a group rack to teach an agent how to do that job or task." This is likened to training a human colleague. This new paradigm is expected to improve AI models and allow individuals to engage in "higher-level things within their companies."

Conclusion and Synthesis

The core argument presented is that while AI is contributing to current job displacement, this is a natural consequence of technological evolution. The more significant takeaway is that AI is simultaneously creating new, high-value job categories, particularly in the realm of training and managing AI systems. The company Merck Core is actively participating in this transition by employing humans to train AI, aiming to enhance productivity and unlock new possibilities, much like previous technological revolutions. The substantial investment in AI infrastructure underscores the belief in its transformative potential for both businesses and society.