Voices from X

Brivael Le Pogam argues that socialism cannot deliver the common good: incentives, dispersed knowledge, and market prices make central planning fail, while historical data from North Korea to Venezuela show catastrophe—not good intentions gone wrong.

https://x.com/brivael/status/2083933065458405379

Google AI Studio introduces Gemini Robotics ER 2, a step change in video understanding, task orchestration, and multi-robot collaboration that brings faster spatial reasoning and more capable physical-world robots.

https://x.com/GoogleAIStudio/status/2082846204262531310

Stefan. decompiles six ordinary moments from a single day — unlocking a phone, cooling coffee, leaning on a table, standing in sunlight, eating lunch, looking at the sky — to show how relativity, entropy, quantum rules, stellar fusion, and light delay run constantly through everyday life.

https://x.com/paradeevic/status/2080803031533170853

Dr Jordan B Peterson argues that Britain's so-called grooming gangs are better understood as sexual conquest gangs: organized networks that systematically raped and degraded vulnerable girls while authorities covered it up to protect multiculturalism.

https://x.com/jordanbpeterson/status/2079970018914381974

Brivael Le Pogam argues that accusations of algorithmic manipulation on X collapse under the open-sourced recommendation code: the real panic is loss of narrative monopoly, not opacity.

https://x.com/brivael/status/2078940564792426577

Austin Koring of ARK argues Meta is already one of AI's biggest winners—not through a new product category, but via compounding AI gains inside its advertising auction that lift conversions across 3.56 billion daily users.

https://x.com/AustinKoringARK/status/2078213465781817670

Clifford Sosin argues that superintelligence arrived incrementally and that intelligence was overrated — the real limit is contact with reality, not pure reasoning power.

https://x.com/CliffordSosin/status/2078594661359194500

Unmesh Joshi argues that domain-specific languages and shared semantic models give LLMs clear boundaries so they generate reliable code—illustrated with PlantUML presentation tooling and the Tickloom framework for distributed systems.

https://martinfowler.com/articles/llm-and-dsls.html

Demis Hassabis argues AGI is only a few short years away and proposes a US-led Frontier AI Standards Body—modeled on a public-private partnership like FINRA—to rigorously test frontier models for national-security risks while still supporting innovation.

https://x.com/demishassabis/status/2076957440109625718

Satya Nadella describes the Reverse Information Paradox of the AI age: enterprises pay twice for intelligence—once with money, and again with proprietary knowledge that leaks through prompts, corrections, and evals—and argues firms need a real trust boundary so their learning compounds inside their own walls.

https://x.com/satyanadella/status/2076323181154230284