AI is not destiny. It is a tool, and a tool magnifies the intention of whoever holds it.

Right now, three hands are at work.

The first uses AI to expand what people can do: faster research, sharper diagnosis, fewer hours spent on drudgery. Done well, that points toward abundance that can be shared.

The second uses AI to cut costs. Headcount becomes a line item to optimize. The savings go to the balance sheet, while the disruption falls on families, communities, and eventually demand itself. Efficiency without distribution leaves trouble to grow.

The third uses AI for war: faster targeting, autonomous weapons, escalation at machine speed. On this path, losses are not shared unevenly. If it wins, everyone loses.

The tool can be the same. The purposes differ.

The question for this decade is not what AI can do. It is what we will decide AI is for. Build tools that make people more capable, not more expendable. Choose deliberately, and choose early.