Does artificial intelligence pose an existential threat?

While claims and open letters about artificial intelligence developing superintelligence and bringing about the end of humanity come to the fore once again, experts point out that real risks stem from those who control the technology rather than the technology itself.

AI Fears and Sci-Fi Scenarios

During periods when leading artificial intelligence researchers sound the alarm and sign open letters, these narratives once again dominate the headlines.

This situation recalls the familiar scenarios of science fiction movies in which artificial intelligence develops superintelligence and destroys everything.

Current Risks and Calls for Regulation

The timing of these claims coincides with public scrutiny over the environmental impacts of data centers and the harms caused to young people by dangerous artificial intelligence companions.

Real risks are embodied by the individuals and organizations that develop and control these technologies, rather than by the technology itself.

Control and Oversight Challenges

Discussions surrounding former Anthropic researcher Jacob Coxon's warning highlight solutions such as safety measures, stronger regulations, and emergency stop buttons.

However, intelligence is not the critical threshold; the real question is whether the system develops its own goals, preferences, or values.

Limits of Rule-Based Control

When a machine begins to interpret rules rather than follow them, rule-based control mechanisms become void.

It is emphasized that in order to properly regulate artificial intelligence, the rhetoric that makes it fascinating needs to be changed.