Why do we assume anyone will give us access to superintelligence?
热度趋势
趋势数据积累中
百分比基于当前可用热度信号，而非评论数或独立用户人数。
一位Reddit用户质疑了普通人将获得超级智能访问权限的假设，即使有公司开发出这种技术。他们提出了一个思想实验，设想一个超级智能AI，如同“被奴役的神”，不太可能通过每月200美元的订阅服务提供。该用户思考在AI历史中，普通人是否有一个有限的窗口期可以相对开放地访问接近前沿的智能，并暗示这个窗口最终可能会关闭。
Thought experiment. Let's suppose a company develops something genuinely superintelligent. The full "enslaved god" scenario, in Tegmark's framework. In what universe would its rational next move be to put it behind a $200/month subscription?
If it can accelerate its own R&D, improve its successor models, discover technologies, protect its infrastructure, make better strategic decisions and increase the organization's advantage, then giving competitors and the general public comparable access is utterly irrational, economically speaking. What am I missing?
Is there a window in AI history during which ordinary people have relatively open access to near-frontier intelligence...and might that window eventually close from either direction?