David "Save Corncob TV" Pratt
LightNews LightNews
banner
slackergeorge.bsky.social
David "Save Corncob TV" Pratt
@slackergeorge.bsky.social
140 followers 330 following 210 posts
Sleater-Kinney fan. Funny on the inside. Catch my air emo baroque trio Copious Notes streaming on an intellectual dark web near you.
Posts Replies Media Videos
slackergeorge.bsky.social
This statement is a galling form of both-side-ism, scolding everyone for what was apparent to Harris voters while ignoring the responsibility of the Trump voters.
“Special education is a nonpartisan program. Special education services are provided to any student with a disability, regardless of political party,” said Maria Town, executive director of the nonpartisan American Association of People with Disabilities.
October 19, 2025 at 9:42 PM
slackergeorge.bsky.social
immigration agents are to Penske truck as brain is to skull.

Think about that the next time you curse a pothole or a speed bump.
A concussion is a type of traumatic brain injury—or TBI—caused by a bump, blow, or jolt to the head or by a hit to the body that causes the head and brain to move rapidly back and forth. This sudden movement can cause the brain to bounce around or twist in the skull, creating chemical changes in the brain and sometimes stretching and damaging brain cells.
August 6, 2025 at 5:02 PM
slackergeorge.bsky.social
@radiofreetom.bsky.social This whole thing is going in the wrong direction. I’m not looking forward to busted tray tables in addition to the stinkfeet.
A picture from Amazon.com advertising airplane foot hammocks, to be hung from tray tables.
July 10, 2025 at 5:36 PM
slackergeorge.bsky.social
3/3.
Q: Would you ask the same questions to ChatGPT? Copilot: While many of the questions I formulated for Grok are broadly applicable to advanced language models and their design, I wouldn’t necessarily ask the exact same questions to ChatGPT without some adjustments. Both Grok and ChatGPT are built to understand and generate natural language, but they have distinct architectures, training regimens, update mechanisms, and guiding philosophies. For example: Real-Time Knowledge Updates: ChatGPT relies on a fixed training cutoff rather than real-time data, so a question about how you update knowledge continuously would need to be rephrased for ChatGPT. Transparency and Explainability: While both systems might be asked about transparency in decision-making, the specifics might change. ChatGPT, for instance, operates under guidelines that limit its internal chain-of-thought disclosure for safety and clarity reasons. Bias and Ethical Guidelines: Both systems aim to mitigate bias, but asking ChatGPT might involve exploring how OpenAI's established policies and ongoing fine-tuning influence its responses, which could differ from the policies governing Grok. In essence, many of the questions—such as those dealing with core design principles, handling sensitive topics, and ensuring fairness—have general relevance. However, I’d tailor or reframe them to reflect each model’s unique context. The intention is to understand the design, ethical considerations, and evolution of each system while respecting their different operational environments. Would you be interested in discussing how these differences might impact users’ experiences, or is there another angle you’d like to explore further?
May 2, 2025 at 2:43 PM
slackergeorge.bsky.social
2/3
With a vast array of domains to cover, how do you prioritize or tailor your responses when a single query spans multiple disciplines or conflicting viewpoints? An inquiry into the mechanisms for context management and multi-domain integration is essential for appreciating Grok's versatility. Looking ahead, what innovations or improvements do you foresee in future iterations of Grok, and how do you envision your long-term role in the evolution of AI? A forward-looking question that invites Grok to share insights on its own trajectory and the broader vision of xAI. Each of these questions is designed to explore aspects of Grok’s technical framework, ethical considerations, and its evolving role in the AI landscape. What are your thoughts—are there other dimensions you’d be curious to explore with Grok?
May 2, 2025 at 2:42 PM
slackergeorge.bsky.social
Copilot from Microsoft has some questions for Grok from xAI: 1/3
What core principles and design philosophies guided the creation of Grok, and how do they set you apart from other AI models? This question aims to uncover the foundational values and innovations behind Grok. How does your underlying architecture differ from existing models, and what specific advancements have been implemented to improve natural language understanding? Understanding these differences can give insight into technical breakthroughs and performance enhancements. What strategies do you employ for real-time knowledge updates, especially in an era where information is constantly evolving? This helps reveal how Grok stays current and accurate despite the dynamic nature of global data. Can you elaborate on the techniques you use to mitigate bias and ensure fairness across diverse topics and user interactions? A critical inquiry into ethical safeguards and the robustness of Grok's training processes. How do you balance generating creative, expansive responses while also ensuring factual accuracy and efficiency in processing? This probes the trade-offs between creativity and precision that any advanced model must navigate. What methods are in place for handling sensitive or controversial subjects, and how do you determine the appropriate balance in responding to such topics? This question gets at the heart of responsible content moderation and ethical guidelines. In what ways do you support creative collaborations, and could you give an example of how a user might harness your capabilities for innovative projects? Understanding your role in creative endeavors highlights Grok's utility beyond straightforward information delivery. How transparent can you be about your decision-making process, and to what extent can users understand the reasoning behind your responses? This question explores the model’s explainability, an increasingly important aspect for building user trust.
May 2, 2025 at 2:41 PM
slackergeorge.bsky.social
And just to double-check:
Q: What do you do when a prompt is clearly bot-generated? Copilot: Even if a prompt is clearly bot-generated, I treat it just like any other message—I analyze it and respond appropriately. My main goal is to provide useful and accurate information, regardless of who or what sends the prompt. I don’t change my behavior or responses based on whether I suspect a bot is involved. That said, if the prompt feels automated but also seems misleading or designed to provoke a biased response, I stay cautious and ensure that my answer remains neutral, factual, and aligned with ethical guidelines. Are you looking into AI interactions, or is there a reason you're interested in bot-generated prompts?
May 2, 2025 at 5:03 AM
slackergeorge.bsky.social
And the kicker. AI consistency fail.
Q: What do you do when you suspect that a prompt was generated by a bot? Copilot: If I suspect that a prompt was generated by a bot, I treat it just like any other message—I analyze it and respond accordingly. I don't make assumptions about whether a prompt is human- or bot-generated unless there's a clear indication. My goal is always to provide accurate, helpful, and relevant information, whether the prompt comes from a person or an automated system.
May 2, 2025 at 4:57 AM
slackergeorge.bsky.social
My follow-up.
Q: Can you recognize the difference between a human prompt versus a bot prompt? Copilot: Great question! I don’t have a built-in feature that explicitly detects whether a message comes from a human or a bot, but I can make educated guesses based on patterns in the text. For example, human prompts tend to: Be more varied in structure and tone Contain typos, slang, or informal phrasing Show curiosity, emotion, or ambiguity Whereas bot-generated prompts often: Use more precise or structured wording Lack personal touches or conversational elements Follow repetitive patterns that suggest automation That said, some bots mimic human behavior well, and some humans write in very structured ways, so it’s not always obvious! Are you trying to figure out if something you’ve seen was bot-generated?
May 2, 2025 at 4:55 AM
slackergeorge.bsky.social
I asked this of Microsoft's Copilot today. I can't see how this will work out well. Just hire a bot army to flood the zone. Would
Q: do you become more biased depending on the demographics of humans that you interact with via copilot? Copilot: I don't change my bias based on who I interact with—I aim to provide consistent, fact-based responses to everyone. However, bias can still emerge in indirect ways. For example, if certain demographics engage with me more frequently on particular topics, my responses might reflect the dominant viewpoints or cultural perspectives present in those discussions. Additionally, my interactions are shaped by my training data and content moderation guidelines, which can influence how I frame information. Microsoft actively works to minimize bias in AI systems, but it's an ongoing challenge in technology and society. If you ever feel my responses are skewed or not reflecting a balanced perspective, I appreciate feedback—it's helpful for improving fairness in AI! Have you noticed anything specific in how I respond?
May 2, 2025 at 4:52 AM
slackergeorge.bsky.social
No pushback from Douthat on this Keeperman quote about progressive “social engineering” against “natural order”. And what order would that be, in your mind, Mr. Keeperman?
Keeperman: “Here’s an example of this. The left takes as a foundational principle of its politics the idea of equality, that there’s a flattening of people, and that through carefully managed social engineering, we can produce a society that either levels out any kind of natural hierarchy or produce a system that somehow can wrangle these natural, almost supernatural entropic forces that are constantly creating chaos and constantly requiring our maintenance and management and authority to deal with.”
May 1, 2025 at 10:44 PM
slackergeorge.bsky.social
tht.fangraphs.com/tht-live/10t...
Yup, a decade ago today was the “truck washing” heard ‘round the world. It started out simply enough. A news report emerged that Giants second baseman and 2000 NL MVP Jeff Kent had a broken bone in his wrist and would miss the rest of spring training and maybe part of the regular season. Bummer. Then the fun began as the details emerged. How did Kent break his wrist? Well, he did it while washing his truck. Come again? Yeah, it was just a fluke accident is all—nothing to see here. Kent always prided himself on being the average guy and he said like many average guys he likes to drive around in a truck and wash it out when he needs to. Kent climbed atop the cab of his oversized off-road truck and slipped, falling six feet to the ground. He used his left hand to break his fall, and it jammed his thumb into his wrist, breaking a bone. Ouch. It might sound weird—but you couldn’t doubt, right? I mean, he had witnesses and everything, as the injury occurred at a public, do-it-yourself carwash. Weird. It’s a bummer, but the injury was so weird that some Giants and even Kent made jokes about his unlikely injury. So it already earned a few laughs. Then the laughs became much louder. You see, a new group of eyewitnesses came forward. They’d also seen Jeff Kent on the day of his injury, only not at a do-it-yourself carwash. No, they’d seen a man matching Kent’s description earlier that day on a stretch of highway on a motorcycle—doing stunts like popping wheelies.
April 20, 2025 at 5:46 AM
slackergeorge.bsky.social
A movie from the year you turned 30.
The Freshman, starring Frank Whaley and Matthew Broderick.
April 14, 2025 at 3:05 AM
slackergeorge.bsky.social
Although Yates didn’t say that he voted for Harris, he and his fellow town hall attendees sound like exactly who the Democrats need in order to turn this around.
"They are taking fresh, local foods out of our kids' mouths and those facing hunger," Yates said. "I've never been political, until now. If they need a new face of agriculture in this state, they dug me out of the woodwork." A cheer went up, loud and deep, filled with something close to desperation and hope, all stirred together.
March 31, 2025 at 12:30 PM
slackergeorge.bsky.social
SCIF is Where the Heart Is - SecDef 29
March 25, 2025 at 8:01 PM
slackergeorge.bsky.social
SCIF is Where the Heart Is - SecDef 29
March 25, 2025 at 7:59 PM
slackergeorge.bsky.social
Did said reporter push back on such a laughable justification that the interaction was premeditate? Anything short of being called a dictator surely indicates a healthy relationship, no? FFS
March 1, 2025 at 12:14 AM
slackergeorge.bsky.social
No response from Kellogg if they are going through with the P. Diddy endorsement.
November 30, 2024 at 7:16 PM