Part 5/7:
On Twitter, the most popular response was that we would retain control over AGI, but it would become autonomous. This indicates a belief that autonomy, rather than malevolence, is the primary concern, and raises questions about our capacity to govern advanced AI systems once they operate independently.
Conversely, on YouTube, which tends to have an audience more familiar with AI alignment work, 43% of respondents believed that we will lose control, but that AGI would be benevolent. This suggests a cautious optimism among more informed or engaged viewers, trusting in the possibility of aligned and safe AI development.