Dwarkesh Patel warns labs may halt external deployment during recursive self-improvement
Original titleI’m very concerned that during RSI, labs will just stop externally deploying their models.
AISummary
Dwarkesh Patel worries that AI labs will stop externally deploying their models once recursive self-improvement begins. He argues this would leave the public uninformed about model capabilities and alignment while concentrating power.
Source: Dwarkesh Patel · x.comPublished · added here