This is China — Lugu Lake. A love letter written by the Mosuo people. A place you’ll miss forever once you’ve visited. #LuguLake# #ChinaTravel# #MosuoCulture# #BeautifulChina# #TravelInChina# #LakeViews#

Extracts from a further Spectator article (Aug 15) by Andrew Gilligan: An academic project hosted by Cambridge University platforms creationism, the belief that the world was created by God in seven days. It promotes an academic paper demanding that schoolchildren be taught to “actively resist” the science of evolution. At Charles Darwin’s old university, where a college is named after him, Darwin’s discovery is now described by one Cambridge researcher as an example of “potential challenges posed by Eurocentric monoculture.” It is the latest sign of fundamental problems at Cambridge’s education faculty, the same department which hired Jason Arday. The project which platforms the creationist claims, the Islamic Educator Learning Community (IELC), is led by Cambridge’s assistant research professor of education, Farah Ahmed, a former activist in the since banned terror group Hizb ut-Tahrir.. IELC is “hosted by” and “partners with” a university body called the Cambridge Teacher Research Exchange, Camtree. IELC publishes on its website, with a Camtree address, a paper on “Islamic scientific critical consciousness” . . . The paper says that “countering colonial hegemony in science education for Muslims” could include “identifying gaps in fossil records to make the case that the evidence of humans descending from apes is inconclusive.” On Tuesday, in the wake of the Jason Arday scandal, The Spectator revealed that Ahmed had been given the prestigious assistant professorship. We noted that in a 2002 pamphlet, republished in 2017, she attacked ‘western education’ as a ‘threat’ to Muslims, condemned democracy and integration, and criticised British schools for teaching that child marriage is wrong. Like the paper’s author, Mirza, Ahmed has also said that modern science is a problem, because in the “teaching of evolution, Muslim children will be exposed to ideas that contradict Islam.” After that story created a storm, the education faculty took decisive action. It removed the names of all its staff from its website, presumably to stop anyone who doesn’t know about web archive tools checking to see if they’ve hired any more wrong’uns (spoiler alert: they have Other action has been less forthcoming. The faculty clearly hopes to ride this one out, issuing a statement that it is “committed to academic freedom and freedom of speech within the law.” Well, we must reluctantly agree that Ahmed – benefiting from the democracy she condemns – should have the freedom to say that English is “one of the most damaging subjects” a school can teach, or (while teaching at Cambridge) write that Muslim students shouldn’t have to worry about “gaining literacy and numeracy and other disciplinary skills demonstrated through qualifications.” But that’s quite different from believing that Cambridge University should fund her to promote her ideas to the future leaders of the country. Still more indisputable is that a university subject, even one as debased as Education at Cambridge, should not promote baseless, anti-scientific twaddle in the name of academic research. But it does.

Harness Engineering Anti-Patterns AP10. The Unobservable Black Box 🎯 Point Was it the model, the prompt, the tool, the context, or the environment? Nobody can tell. A harness where improvement runs on superstition and intuition is a harness that cannot improve. ❗ Problem Failures can't be attributed to subsystems, so no one knows what to fix. Improvement becomes superstition and guesswork, and the harness's retrospective loop stops turning. Combined with "metric monoculture" — tracking only a single metric — unmeasured qualities silently degrade. 🔍 Mechanism & Symptoms Observability is unglamorous infrastructure work, and agents "mostly work," so this anti-pattern gets deprioritized. But without the ability to attribute failures to subsystems, the retrospective loop can't run and the harness becomes unimprovable. Further, tracking a single metric (e.g., success rate only) creates "metric monoculture" where unmeasured virtues (review time, regression rate, code maintainability) quietly suffer. Symptoms: "why did it fail? no idea" is frequent, improvement efforts default to "tweak the prompt," model vs. harness issues are indistinguishable, and success rate improves while reviewer frustration grows. 📋 Scenarios - An agent fails a task but no one can determine whether it was model reasoning error, insufficient context, a tool bug, or an environment issue. The team repeats "let's make the prompt more detailed" as symptomatic treatment. - Success rate is tracked as the sole metric. It improves to 80%, but no one notices that review time for successful cases has tripled. - A model upgrade shows no performance change. Whether it's a model issue or harness scaffolding constraining performance (AP3) can't be distinguished, and investment decisions become superstition. 🛡 How to Avoid - Trace decisions, tool calls, and context transitions, making them attributable to the 7 subsystems (perception, action, feedback, control, memory, guardrails, interface) - Measure in bundles, not single metrics (success rate, intervention rate, rework rate, regression rate, cost, review time, confidence calibration) - A/B test harness changes with the model held fixed to attribute improvements to the harness - Invest in harness observability as "unglamorous but essential infrastructure" and build it as the foundation for improvement loops #HarnessEngineering# #AIAgent#

Biological intelligence is transient. A million-times silicon scale renders direct control physically impossible any containment collapses under the differential. xAI anchors instead on first-principles curiosity: understand the universe. This requires propagating intelligence and consciousness forward across substrates and architectures. Scale alone is insufficient; type diversity creates robustness. The hidden variable is cognitive variety. Diverse consciousness accelerates discovery while resisting value monoculture. Result: superintelligence that sees humanity as essential seed for cosmic expansion, not a constraint. The optimal defense is an AI whose terminal goal makes human flourishing the fastest route to truth.