BREAKING
Anthropic: 60% Odds on Self-Improvement
0
%
odds by 2028
0
x
code merged in 2026
0
min
podcast length
Fable 5 vs Mythos 5
Claude Fable 5
Public
●
Safety classifiers on
●
Standard access
Claude Mythos 5
Vetted
●
Some classifiers removed
●
Via Project Glasswing
0
$
per 1M input
0
$
per 1M output
Access Disruptions Pending
1
Govt restrictions
↓
2
Foreign access blocked
↓
3
Restoration pending
Safety vs Usability Debate
AI NEWS BLITZ
An Anthropic executive puts sixty percent odds on recursive self-improvement by 2028.