Announcement
“Poor framing”: @kaicathyc says OpenAI’s recursive self-improvement wording was revised
According to @kaicathyc, revised wording emphasizes preparing safeguards ahead of models capable of fully automated recursive self-improvement.
TLDR
When @_NathanCalvin questioned whether OpenAI’s system card implied plans for fully automated recursive self-improvement, @kaicathyc called the wording “poor framing” and said it was updated. The replacement describes the “High” threshold as “an early milestone for preparing safeguards ahead of models capable of fully automated, recursive self improvement.” The reply does not spell out OpenAI’s research plans.
Combined views
2.9K
3 Sources, first seen ago
Comments