Bill Gates, the former Microsoft founder, recently publicly shared his core views on the long-term development of AI, stating that it's time to consider levying a robot tax at this stage, using policy measures to buffer the social shocks caused by AI completely replacing human labor. He affirmed that the open letter "Pacing the Frontier," signed by top global professionals calling for slowing down cutting-edge AI, is worth considering for policymakers, but he also doubts whether such self-regulation by the industry can truly take effect.
Slowing Down "Machine Replacement" Through Taxation
Gates pointed out the counterproductive incentives in current tax rules: companies must pay payroll taxes when hiring people, but purchasing robots or introducing AI systems can be counted as costs and enjoy deductions, which objectively encourages companies to replace humans with machines. He envisions a special tax on automated systems that replace human labor, which would slow down the uncontrolled elimination of human jobs, and the tax revenue could be specifically used for retraining laid-off workers and universal social security to ease the impact of employment changes.
He also introduced a new concept called "human-only domains": through policy, certain special jobs should be designated as ones that can only or mainly be performed by humans, strictly limiting AI from entering. Gates believes this is easier to implement than most people imagine, essentially preserving jobs for economic stability—without intervention, if AI takes over industries entirely, older workers or those with limited skills will struggle to switch careers, and the government has a responsibility to intervene to avoid upheaval.
The scope of this domain is not fixed and will change dynamically with industrial structure. Countries can first identify priority jobs to retain while opening up other scenarios, leaving several years or even decades as a buffer period, ensuring that AI penetration always maintains enough human jobs, avoiding the full cost of upgrading being borne by ordinary workers. Gates' approach offers a new direction for global AI regulation: technology itself is not the problem; without social security to cushion the impact, technological substitution will only benefit capital while shifting the costs to workers.
