/Tech19h ago

Aria fine-tuner @xlr8harder and researcher Herbie Bradley argue differing model specifications drive contrasting behaviors at OpenAI and Anthropic

The behavioral gaps widen as underlying model capabilities advance.

0901392

#1424

Original post

xlr8harder@xlr8harder#1855inTech

@hamandcheese The companies have different underlying philosophies, and this absolutely informs their product direction. It's becoming more visible now because capabilities are reaching the point where the philosophies dictate more divergent paths.

Samuel Hammond 🦉@hamandcheese

This tweet confuses me insofar as Ant and OpenAI are both building more-or-less the same thing using more-or-less the same paradigm. Whether AIs are sentient and whether RSI fooms to a machine god aren't determined by corporate values statements.

12:30 AM · Jun 12, 2026 · 249 Views

Sentiment

Sentiment building, check back later.

Cluster Engagement

Posts from X

Most Activity

VIEWS145LIKES2

Herbie Bradley@herbiebradley

@hamandcheese I just read this as differences in approaches to writing a model spec/constitution which seem real and likely to end up with different results

Samuel Hammond 🦉@hamandcheese

2h14520