AI should amplify, not replace, human judgment: Report

Artificial intelligence (AI) should not replace human judgment but amplify it, according to a new manifesto released by AI research company Thinking Machines, which argues that the "future worth building is human" and calls for AI systems that evolve alongside people rather than operate independently.

In a blog titled 'The Future Worth Building Is Human', the company outlines its vision for building AI that extends "human will and judgment" by embedding human knowledge, feedback and values into the development of advanced AI systems.

The report contends that while AI capabilities are advancing rapidly, the decisions about what AI should do must remain in human hands.


"We believe the future worth building is human -- shaped by human knowledge, guided by human will, and decided by human judgment," the report states, underscoring the company's central philosophy that AI should function as a collaborator rather than a replacement for people.

According to the report, much of the world's productive knowledge is local, tacit and continuously evolving through human experience. It argues that centralized AI systems trained in limited environments cannot fully capture the expertise generated by individuals across organisations and industries.

"AI exists to serve the work that we do," the report says, adding that the knowledge driving this work is "generated continuously by people engaged in the work."

The company also challenges the prevailing trend of developing increasingly autonomous AI systems that minimise human involvement. Instead, it advocates designing AI models capable of inviting constant human participation through richer interactions, continuous feedback and greater customization.

The report says current AI interfaces remain too limited to fully communicate human intent and expertise, describing today's interaction methods as insufficient for sustained collaboration between humans and machines. It argues that future AI systems should be designed to work with people in real time rather than simply execute instructions autonomously.

Thinking Machines further argues that AI alignment should not be dictated by a handful of organizations. Instead, it proposes what it calls "decentralized alignment," where individuals and organisations can shape AI models according to their own values while maintaining safety.

The report warns that concentrating AI development and value alignment within a few entities risks reducing diversity of thought and creating excessive concentrations of power. It advocates building ecosystems of AI models that reflect different perspectives and evolve through human participation rather than standardization.

Summing up its vision, the company says the choice is not between human dominance and human obsolescence but between different paths of AI development.

"The future is not a choice between human dominance and rapid obsolescence in the face of AI. Different roads lead to many different futures, and we get to choose which one to take. We are building technology that lets the born and the made walk the road together," the report concludes.