Whenever I don’t use codex for a task, I ask myself why and usually realize that there’s some missing context, I needed to write a skill, or I just didn’t think to use it. Rarely is it because the task is outside of the capabilities of the model. Overhang right now feels large.
LLMS
-
LLMs are not conscious, two true versions
By
–
LLMs are not conscious.
LLMs are not not conscious.
Both are true. -
Apple will unveil deep integration of Gemini 1.2t into its OS
By
–
Apple's WWDC starts on Monday. This year it promises to be particularly interesting. The secret will be revealed of how Apple has finally integrated a proper AI model deep into its operating system. As is known, it's a Gemini version, developed specifically for Apple, with 1.2t
-
Self-Trained Verification for Self-Improvement in Training and Test Phases
By
–
Reasoning models improve faster with a good verifier, but verifiers cannot learn to detect subtle errors by themselves. However,
-
Hope for Codex Desktop controlling other desktop instances
By
–
hope we aren’t far from codex desktop controlling other desktop instances
-
Your Photos Cost You, AI Makes Them Professional
By
–
La plupart des propriétaires ne réalisent pas que leurs photos leur coûtent de l'argent.
— Jouhatsu | AI Influence Operator (@Jouhatsu_ai) 6 juin 2026
Cela a commencé comme une simple photo prise avec un téléphone.
Claude + Higgsfield MCP l'ont transformé en quelque chose qui ressemble à une mise en scène professionnelle en quelques… https://t.co/JEDC92DiY4 pic.twitter.com/H5D8N6QOaDMost owners do not realize that their photos cost them money. It started as a simple photo taken with a phone. Claude + Higgsfield MCP transformed it into something that looks like a professional staging in a few…
-
Undetected AI hallucinations become users’ false beliefs.
By
–
The vast majority of AIs' hallucinations go undetected. They just become their users' false beliefs.
-
AI writing quality issues in software menus
By
–
One reason you want AIs to be better writers is that there is a lot of writing even in software, and it is incredibly painful to hit a menu which is filled with Claudisms or ChatGPTish phrases. A report is not "what leaves the room" & analyses are not "every number makes a mark"
-
AI cafe agent given spending authority despite sensible caution
By
–
Most people are sensible enough not to give purchasing authority to anything attached to an LLM This cafe-running agent got to spend money https://
andonlabs.com/blog/ai-cafe-s
tockholm
… -
Proactive-Soul: New OpenClaw Skill for Agent Inner Life
By
–
Just published proactive-soul to ClawHub. This is an @openclaw skill that gives your agent an inner life. Improves the digital mind of your agents in ways that will amaze and make your entire setup more productive for you. Makes dialog with agents more intelligent, but also