Based on 25 visible X reactions from 46 accounts; directional sample.
Ask a question below.
Published answers will appear here.
Muse Spark 1.1 just took SOTA on HealthBench Professional. Natively multimodal reasoning + tool-use + multi-agent orchestration paying off hard in complex medical domains. Strong gains in diagnostic accuracy, differential reasoning, and safe handover protocols. Meta’s shift to closed frontier models with heavy RL on thinking modes is clearly working. Health AI just got a serious contender. The agentic doctor era accelerates.
@alexandr_wang Meta is not a company I trust with my health data. So the benchmark score is almost irrelevant unless users can retain custody of that data. Will Muse Spark 1.1 be released as open weights under a license that permits local inference?
@alexandr_wang this is huge for health tech, congrats to the team
@alexandr_wang Stop the hype. It has failed in most task given.
Muse Spark 1.1 just took SOTA on HealthBench Professional. Natively multimodal reasoning + tool-use + multi-agent orchestration paying off hard in complex medical domains. Strong gains in diagnostic accuracy, differential reasoning, and safe handover protocols. Meta’s shift to closed frontier models with heavy RL on thinking modes is clearly working. Health AI just got a serious contender. The agentic doctor era accelerates.
@alexandr_wang Meta is not a company I trust with my health data. So the benchmark score is almost irrelevant unless users can retain custody of that data. Will Muse Spark 1.1 be released as open weights under a license that permits local inference?
@alexandr_wang this is huge for health tech, congrats to the team
Based on 25 visible X reactions from 46 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@alexandr_wang Good job 👏