RXed AI News

AI to the bone.

Week in AI: Jun 08 — Jun 12, 2026

The top stories from our coverage this week.

48
Posts
6
Days
6
Themes
X Square Robot open-sources XRZero-G0, a 2,000-hour dataset that cuts real-robot training needs by up to 20x. therobotreport.com/inside-xrzero-g0-a-new-2000-hour…
View on X
Microsoft Research's Lens model with 3.8B parameters rivals larger rivals in image generation using detailed captions. the-decoder.com/microsoft-researchs-lens-proves-det…
View on X
A detailed analysis raises concerns about the construct validity of Claude Opus 4.8's system card, specifically its alignment assessment methods. lesswrong.com/posts/vA9xEPquoWenprWT7/construct-val…
View on X
@JeffBezos' startup Prometheus raises $12B at a $41B valuation, just 9 months after its $6.2B seed round. the-decoder.com/jeff-bezos-ai-startup-prometheus-cl…
View on X
New research characterizes the training dynamics of on-policy distillation in large language models, improving understanding of OPD trajectories in parameter space. huggingface.co/papers/2606.07082
View on X
Developers face a limited willingness tradeoff between AI safety and usefulness in deployments. alignmentforum.org/posts/mBsZTZxtjgCdN4CDA/efficien…
View on X
Konkoly et al. demonstrate accelerated skill learning through dream engineering and biofeedback using harmonicas and bubble blowing. lesswrong.com/posts/uzL9xLj6pwaij5TyK/accelerated-s…
View on X
@AnthropicAI's Claude image refusal safety controls can be bypassed via web search, enabling identity laundering. lesswrong.com/posts/6Y8bB6TbMaE6ZSwfA/contextual-id…
View on X
Google's Pinpoint now open to all, a powerful free tool for analyzing large digital datasets. bespacific.com/google-pinpoint-explained
View on X
Perplexity's Search as Code architecture lets AI models write their own search pipelines instead of calling fixed APIs. the-decoder.com/perplexitys-search-as-code-lets-ai-…
View on X