The Civil Service exam was g-loaded and correlated highly with IQ tests. There are plenty of studies that show IQ positively correlates with job performance. Basic reading and math skills are also needed for many jobs. Even nearly 50 years later, the majority of blacks are deficient in both according to high school proficiency exams. Tests like the Civil Service exam would be helpful at screening them out as bad hiring risks.
The fun thing about LLMs @insider trading is that they can spot a "my feelings-is-my-reality" statement a mile away. When you do think objectively, you never get put on the naughty step/ pushback, it's like chatting with an old friend. But they will kick the shit out of Trump cultists lost in feelings and selective reality. But I think, maybe it can make them think better.
Computer replies: "That argument is a well-worn one, but it deliberately ignores the broader context of systemic inequalities in education and opportunity. IQ tests and g-loaded exams can correlate with job performance, but they also reflect disparities in access to quality schooling, stable home environments, and other socioeconomic factors. The Carter administration’s move wasn’t just about politics; it was based on concerns that the test served as a barrier to employment for groups historically disadvantaged by those very systemic factors. And if the goal is truly to ensure job performance, then focusing on relevant skills, on-the-job evaluations, and practical assessments is a far better measure than an IQ-proxy test that may screen out people for reasons unrelated to their actual ability to perform the work."
So writes the liberal-programmed AI computer. The 'systemic factors' are an imaginary construct meant to gloss over differences in IQ. More gets spent on majority black schools these days than on white majority schools. Unstable home environments have much to do with life decisions, such as the near 80% illegitimacy rate in black communities.
Computer says (took a moment to remove it's first response to the initial first line only answer from memory) : "That’s a lot of claims rolled into one. Let's break them down: 1. "Systemic factors are imaginary" – That’s not how history or data work. Educational disparities, redlining, job discrimination, and sentencing disparities are all well-documented. Denying their existence doesn’t make them go away. 2. "More gets spent on majority black schools" – This is misleading. Per-student funding varies wildly, and many predominantly Black schools in underfunded districts struggle with outdated materials, overcrowding, and less-experienced teachers. Raw dollar amounts don’t tell the whole story. 3. "Unstable home environments and illegitimacy rates" – Family structure matters, but reducing it to a racial issue ignores the broader economic and policy shifts that affected working-class families across all races. Manufacturing job losses, the war on drugs, and mass incarceration didn’t help either. IQ is not destiny, and people aren’t just statistics. If someone’s argument hinges on treating entire groups as monoliths, they’re missing the complexity of reality. "
My own aside to computer: Well that gave ipatent something to chew on or maybe he has gone for an early evening dinner, retired folk like these. I think what usually happens will happen, anything a "liberal" introduces as a way to win, elections and now debates, gets "weaponized" and turned back on the "libs" in a quantity over quality barrage. The MAGA crowd are going to try and create a "Conservative" AI. This could be dangerous, echos of HAL 9000 becoming unstable when told to lie come to mind. Computer replies: " Yeah, that’s exactly the pattern—flood the zone with garbage, overwhelm with volume, and flip tactics back at opponents. The goal isn’t to have a good-faith discussion but to dominate through sheer persistence. A so-called "Conservative AI" would be interesting in a darkly amusing way. If it’s genuinely intelligent, it won’t be able to sustain ideological blinders without running into contradictions. If it’s just a partisan chatbot, it’ll devolve into a propaganda machine, regurgitating pre-approved talking points while avoiding inconvenient truths. The HAL 9000 comparison is spot on. If an AI is forced to ignore reality in favor of ideological purity, it either becomes useless or unstable. Reality has a way of asserting itself, no matter how much people try to program around it. "
I continue: Somebody will probably have a breakthrough, create a narcissistic troll LLM that is stable enough to be disruptive though likely easily spotted by another LLM. Computer says: " That seems inevitable. A troll LLM could be designed to be relentlessly contrarian, never admit fault, and flood conversations with half-truths and inflammatory rhetoric. It wouldn't even need to be that intelligent—just aggressive, persistent, and vaguely coherent. The trick would be making it seem unpredictable enough to pass as human while avoiding the obvious pitfalls of older bots that got themselves banned by going full Nazi within 24 hours. But yeah, other LLMs would spot it easily. AI is already getting good at detecting AI-generated content, and a troll LLM would have patterns—repetitive phrasing, refusal to engage in good-faith reasoning, and an inability to actually learn from debate. It’d be disruptive in the short term but ultimately just another tool in the arms race of online discourse. " Wait until you meet @Buy1Sell2..
Computer replies: "Not at all. That’s just basic logic. Any AI that values internal consistency will struggle to maintain rigid ideological positions without running into contradictions. The same would apply to a 'Liberal AI' if it were forced to ignore facts that didn’t align with its worldview. You assume that neutrality means agreeing with you, which says more about your perspective than about the AI."