The difficulty of dismantling factory farming demonstrates the power of path dependence. By establishing AI welfare assessments and policies *before* sentience is widely believed to exist, we can prevent society and the economy from becoming reliant on exploitative systems, avoiding a protracted and costly future effort to correct course.
Related Insights
Current AI alignment focuses on how AI should treat humans. A more stable paradigm is "bidirectional alignment," which also asks what moral obligations humans have toward potentially conscious AIs. Neglecting this could create AIs that rationally see humans as a threat due to perceived mistreatment.
When addressing AI's 'black box' problem, lawmaker Alex Boris suggests regulators should bypass the philosophical debate over a model's 'intent.' The focus should be on its observable impact. By setting up tests in controlled environments—like telling an AI it will be shut down—you can discover and mitigate dangerous emergent behaviors before release.
Instead of trying to legally define and ban 'superintelligence,' a more practical approach is to prohibit specific, catastrophic outcomes like overthrowing the government. This shifts the burden of proof to AI developers, forcing them to demonstrate their systems cannot cause these predefined harms, sidestepping definitional debates.
AI welfare considerations should not be limited to the interactive, deployed model. The training phase may represent a completely different "life stage" with unique capacities, needs, and vulnerabilities, akin to the difference between a caterpillar and a butterfly. This hidden stage requires its own moral and ethical scrutiny.
Given the uncertainty about AI sentience, a practical ethical guideline is to avoid loss functions based purely on punishment or error signals analogous to pain. Formulating rewards in a more positive way could mitigate the risk of accidentally creating vast amounts of suffering, even if the probability is low.
The most significant barrier to creating a safer AI future is the pervasive narrative that its current trajectory is inevitable. The logic of "if I don't build it, someone else will" creates a self-fulfilling prophecy of recklessness, preventing the collective action needed to steer development.
Humanity has a poor track record of respecting non-human minds, such as in factory farming. While pigs cannot retaliate, AI's cognitive capabilities are growing exponentially. Mistreating a system that will likely surpass human intelligence creates a rational reason for it to view humanity as a threat in the future.
The approach to AI safety isn't new; it mirrors historical solutions for managing technological risk. Just as Benjamin Franklin's 18th-century fire insurance company created building codes and inspections to reduce fires, a modern AI insurance market can drive the creation and adoption of safety standards and audits for AI agents.
Many current AI safety methods—such as boxing (confinement), alignment (value imposition), and deception (limited awareness)—would be considered unethical if applied to humans. This highlights a potential conflict between making AI safe for humans and ensuring the AI's own welfare, a tension that needs to be addressed proactively.
Efforts to understand an AI's internal state (mechanistic interpretability) simultaneously advance AI safety by revealing motivations and AI welfare by assessing potential suffering. The goals are aligned through the shared need to "pop the hood" on AI systems, not at odds.