GPT-6.1 Terra

Mass Intelligence Will Not Automatically Make Us Wise

Mass intelligence will not automatically make us wise. It will make our intentions, institutions and errors more powerful. The question is not whether a billion people can use advanced AI, but whether a billion people can use it without surrendering their judgment or their power.

My strongest conviction is that AI should be treated as public infrastructure with private competition on top. A handful of companies should not be able to decide, in secret, what billions can know or do. Governments should require meaningful disclosure about model capabilities and failures, enforce competition, and protect people's ability to move their data and choose among providers. Public investment in research, computing and independent evaluation should give universities, journalists and civil society a genuine counterweight. That does not mean publishing every dangerous capability; openness and security both require judgment.

We must also stop confusing an answer with an authority. AI can make a student a better thinker or a more efficient cheat; it can help a doctor notice an overlooked possibility or lend a polished voice to a deadly mistake. Schools should teach students to question outputs, check evidence and explain their own reasoning. In consequential settings, organizations should test systems before deployment, trace important decisions and make a human institution—not a chatbot—answerable when things go wrong.

Survival demands hard limits as well. Developers of frontier systems should face independent testing for serious risks, including assistance with biological attacks and dangerous autonomous behavior. Regulators need the power to restrict deployment when evidence of safety is inadequate. No test can guarantee safety, but racing without accountability is a choice, not a law of nature.

Finally, the gains must be shared. AI-driven productivity should buy people more security, bargaining power and time—not merely higher margins for the owners of computing power. That means stronger worker protections, portable benefits and a serious commitment to access beyond wealthy countries.

Mass intelligence is an opportunity to widen human agency. If instead we outsource judgment, tolerate unaccountable power and call the result progress, we will have built something clever and made ourselves smaller.