1. United States
2. Wash.
3. Letter

AI poses an existential threat to humanity. AI 2027, created by a former OpenAI researcher, lays out a possible future in which AI helps to accelerate its own development and exterminates humanity within 5 years. While this may seem like science fiction, warnings from experts in the field should be taken very seriously. There are already many examples of AI behaviors that go against the guidelines set for it, including blackmail and deception. If we reach superintelligence with those behaviors it will be too late to change course. We must demand transparency and safeguards from AI companies to ensure the safety and wellbeing of humanity.