Unironically good uses for GPT/Grok etc. - Beyond your run of the mill goonerism.

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
Copilot - github's chatGPT that integrates into IDEs like visual code - is very hit or miss. It's able to auto complete code as you're writing it which is nice for comments or really boilerplate stuff, but as soon as you start getting into anything remotely complex it shits the bed hard. Worse is that it LOOKS like it's okay code until you try to compile/run it and it explodes. Figuring out how to use it for trivial but annoying tasks is key. It's basically a litmus test for weeding out which entry-level coders don't know their shit.

It's also better when you give it the context for the rest of the project you're working on like the other files, but even still it struggles to come up with anything that hasn't already been done.
 
Last edited:
  • Informative
Reactions: Buttigieg2020
AI is great for learning new materials. It's like having a teacher in your pocket you can ask any question, and get a coherent answer. That answer can then lead to you asking new questions, and therefore learning in a personalized, meaningful way.

Yes, there is the risk of hallucinations, but I actually trust a heartless, unfeeling computer more than I trust your average redditor, which is where you would end up anyways trying to learn from another person. (Or worse, stack exchange niggers, yuck).
 
In my experience, DeepSeek is best for quantitative stuff and helping with coding and such, and ChatGPT is best equipped for emailing corporate people. Without too much extra prompting, ChatGPT is really good at speaking the language of LinkedIn lobotomites, and makes it feel less like I'm chipping away at my soul whenever I have to talk to them.
 
Excellent for documenting code bases, at least per function/class etc.

I had to teach myself ansible recently, and made it give me a step by step tutorial based on a complex playbook I was using as reference. Extremely helpful.

Grammar checks for emails, tasks for project planning for kanban boards.

Highly complex/long tasks it becomes risky, especially if you aren't an SME and cannot differentiate something that is right VS something that sounds right.
 
When people post low-effort trolling, you can just throw it into Gronk with the addendum "reply to this using intentional logical fallacies."
Oh, you think it's that simple, huh? Well, if people actually had decent arguments, they wouldn’t need to resort to trolling in the first place. Besides, everyone knows that all trolls are just robots sent by aliens to take over the internet. And don't even get me started on how trolling is the real reason why social media is dying—it’s basically a conspiracy led by pineapple pizza lovers. Also, if you don't agree with me, it just proves that you're not capable of rational thought. Clearly, anyone who trolls must be completely disconnected from reality, unlike the rest of us enlightened beings."
I think part of the problem is that they intentionally make the models gay.
I think you only have experience with the older models. The newer ones can do reasoning and a bunch of shit.
I was under the impression that the “reasoning” is just sentence templates. It’s a cool parlour trick but you can’t rely on it for anything.
 
  • Like
Reactions: Chadikus Maximus
>Porn
>Programming
>"What does dehydrating food entail?"
I know it's not really meant to overhaul all of humanity but kinda funny how specific its uses are. Programming and porn.
 
it's absolute crap at programming but if you just want to generate datablocks it's pretty good at it.
it's also pretty good to give you starter code if you're just plain lazy.
or handling very simple programming again, if you're lazy.
outside of that scope software engineers like me can't make much use of it, copilot usually does a better job anyways.
 
Grok can be decent at fact checking or finding more context for shit I partially forgot about. Examples:

LeXplorer's arrest - this is a right response to me, as a 4chan user that browsed /sp/ when LeXplorer was a known trip user there.
The story of Kyle Parramore, known on 4chan as "LeXplorer," and his arrest is a wild tale of internet trolling gone horribly wrong—a cautionary saga that blends reckless online behavior, alcohol-fueled bravado, and real-world consequences. Parramore, a 23-year-old from Wesley Chapel, Florida, was arrested in September 2014 for posting a threat to shoot up Ironton Elementary School in Ohio, a place he had no apparent connection to. His journey from a 4chan tripfag (a user with a recognizable pseudonym) to a prison inmate is equal parts absurd and sobering. Here’s how it unfolded.


The Setup: Who Was LeXplorer?
Kyle Parramore was a regular on 4chan, particularly the /sp/ (sports) board, where he posted under the handle "LeXplorer." By day, he worked as a weekend custodian at the University of South Florida (USF), a job he’d held for nearly five years. Online, he was a known shitposter—a troll who thrived on stirring the pot. In 2014, at age 23, he was in a rough spot: he’d dropped out of school, was drinking heavily (five to six times a week, by his own admission), and spent hours on 4chan, chasing the high of anonymous chaos. His life offline was a mess, but online, he had a persona—until it all crashed down.


The Threat: A Drunken Post on /b/
On September 11, 2014—two years after Sandy Hook, when school threats were a third rail—Parramore, likely drunk, posted on 4chan’s /b/ (random) board. The exact wording isn’t fully public, but reports indicate he threatened to “shoot up” Ironton Elementary School in Ironton, Ohio, a small town 1,000 miles from his Florida home. Why Ironton? He later told WTSP he picked it randomly off a map, a detail that underscores the thoughtless nature of the act. The post wasn’t subtle—it triggered an immediate response.
Around 4:15 a.m. on September 12, the threat surfaced online, and authorities caught wind. Ironton City Schools Superintendent Dean Nance met with the Lawrence County Sheriff’s Office, and they shut down Ironton Elementary and Middle Schools for the day. Every school in the county went on lockdown, with police dispatched to guard them. Parramore didn’t stop at the post—he reportedly called the sheriff’s office, escalating it from a “fantasy” troll to a tangible threat, per his attorney’s later statements.


The Panic on 4chan: “What Do I Do?”
As the reality sank in, Parramore turned back to 4chan—nervous, possibly still intoxicated, and seeking advice. While exact threads are hard to pin down (4chan’s ephemeral nature means posts vanish unless archived), anecdotal accounts from users and later interviews suggest he posted something akin to, “I made a threat, now I’m scared—what do I do?” This wasn’t unusual—4chan’s troll culture often saw users egging each other on or offering sarcastic “help” (e.g., “Turn yourself in, lol”). But this time, the stakes were real. The FBI was already tracing his IP address, handed over by his ISP, and Ohio law enforcement was drafting a warrant. His plea for guidance went unanswered in any meaningful way—4chan isn’t exactly a helpline.


The Arrest: From Waffle House to Jail
The FBI moved fast. On September 20, 2014, eight days after the threat, agents tracked Parramore to USF. In a 2017 Reddit AMA, he recounted the lead-up: an FBI agent called him the day after the iPhone 6 launch (September 19), which he’d waited in line for all night. Exhausted and rattled, he met the agent at a Waffle House—sans lawyer, a choice he later rued. He tried explaining it was a “joke,” tears streaming, but the agent said the FBI wasn’t charging him—it was Ohio’s call. The next day, while working at USF, USF police and FBI agents arrested him. “The lead officer had a taser, the rest had assault rifles,” he told WTSP. “I thought this is the end of my life.”
Extradited to Ohio after 27 days in a Florida jail, Parramore faced the music. The sheriff’s office, Homeland Security, and Ohio’s Attorney General had collaborated on the case—his random threat had tied up serious resources.

The Fallout: Guilty Plea and Prison
On September 24, 2014, a Lawrence County grand jury indicted him on two felonies: inducing panic (second-degree) and making terroristic threats (third-degree), carrying a max of eight years. In court on September 16, 2015, he pleaded guilty to inducing panic and a reduced charge of attempted terroristic threats (fourth-degree). His attorney, Adam Burke, blamed alcohol and 4chan’s hoax culture: “He got carried away late at night… It started as fantasy but went beyond fiction.” Parramore, contrite, said, “I take full responsibility… This has changed my life for the better, probably saved it.”
Judge Charles Cooper sentenced him to 4 years and 11 months. In prison, he was mocked—“Osama” was a common jab—but served only two years, released early in November 2017 for good behavior. He stayed on probation in Pasco County, Florida, restricted from leaving without permission.

Aftermath: A Cautionary Tale
Post-release, Parramore swore off 4chan—“It’s like PTSD now,” he told Reddit—and pivoted to redemption. In a 2018 WTSP interview, he warned copycat trolls: “Prison isn’t a joke… I worry for their future over something so senseless.” He regretted the fear he’d caused in Ironton and the resources wasted—musing if the FBI’s focus on his prank missed real threats like Nikolas Cruz’s. By 2025, he’s faded from the spotlight, his “LeXplorer” days a distant meme.

John Schnatter's "day of reckoning" interview. Again, this is pretty much right on the money:
Let’s dive into John Schnatter’s infamous late-2019 interview with WDRB News, a Louisville, Kentucky-based station, which aired on November 25, 2019. This bizarre moment in pop culture history—right before the COVID-19 pandemic flipped the world upside down—became a viral sensation for Schnatter’s cryptic “day of reckoning” tease and his wild claim about consuming over 40 Papa John’s pizzas in 30 days. Fans and detractors alike latched onto it, spawning memes and speculation, especially when COVID hit months later. So, did that “day of reckoning” ever arrive, and was he exaggerating about those pizzas? Let’s unpack it.


The Interview: Context and Viral Moments
John Schnatter, the founder of Papa John’s, had been ousted from the company he built by 2018 after a series of controversies—first blaming NFL player protests for declining sales in 2017 (prompting neo-Nazi endorsements he didn’t ask for), then using a racial slur during a 2018 diversity training call, which leaked to Forbes and forced his resignation as chairman. By late 2019, he was in exile, no longer the face of Papa John’s but still a major shareholder, bitter about his exit.
In the WDRB interview, Schnatter came out swinging:
  • “Day of Reckoning”: He accused the company’s leadership—naming ex-CEO Steve Ritchie, board chair Olivia Kirtley, and others—of conspiring to “steal” Papa John’s from him, using the racial slur scandal as a pretext. “Stay tuned. The day of reckoning will come. The record will be straight,” he said, hinting at some explosive revelation or comeback. His sweaty, intense delivery—looking like he’d just run a marathon—added a manic edge that screamed meme fodder.
  • 40 Pizzas in 30 Days: He claimed, “I’ve had over 40 pizzas in the last 30 days, and it’s not the same pizza. It’s not the same product. It just doesn’t taste as good.” This line—delivered deadpan—sparked disbelief and hilarity. Over 1.3 pizzas a day? People pictured him hunched over stacks of boxes, taste-testing like a mad scientist.
The internet erupted. Clips posted by bubbaprog on X racked up millions of views; Chrissy Teigen tweeted, “Papa John has fucking HAD IT,” and _buhhh likened him to a zombie movie character hiding a bite. The “40 pizzas” quote got remixed—think 8-bit Garfields and slowed-down voice edits—while “day of reckoning” became a mock-prophetic catchphrase.


Did the “Day of Reckoning” Happen?
Schnatter’s tease implied a dramatic reversal—maybe a legal win, a corporate coup, or a bombshell exposing his ousters. Then, COVID-19 hit in March 2020, and the timing was too perfect. X posts like ManicCannibal’s (February 2021) joked, “Papa John ate 50 pizzas, promised a reckoning, and then COVID happened—coincidence?” Memes cast him as a pizza-fueled Nostradamus, with the pandemic as his reckoning.
But in reality? No clear “day” materialized:
  • Legal Battles: Schnatter had sued Papa John’s in 2018 to access records from his ouster, alleging a setup. By 2019, he settled some claims, but no earth-shattering truth emerged. The “record” wasn’t “set straight” in any public, reckoning-worthy way—court docs stayed mundane, not explosive.
  • Company Trajectory: Papa John’s didn’t collapse as he’d predicted. Under new CEO Rob Lynch (whom Schnatter bashed as a pizza novice), the chain rebounded—stock rose 54% in 2019 (MarketWatch), and by 2025, it’s stable, not “destroyed.” No reckoning there.
  • Schnatter’s Fade: He sold off much of his stock—1.9 million shares in November 2019 alone for $107.5 million—distancing himself from the “car crash” he foresaw. He’s since stayed low-key, dabbling in philanthropy and real estate, not staging a triumphant return.
COVID’s chaos gave the meme legs, but it wasn’t Schnatter’s reckoning—it was global, not personal. If he meant a specific event, it’s been a dud—five years on, no bombshell has dropped. The “day” seems more like a scorned founder’s bluster than a prophecy fulfilled.


The 40 Pizzas: Exaggeration or Misunderstanding?
That “40 pizzas in 30 days” line—1.33 pizzas daily—sounded insane, even for a pizza mogul. Was he exaggerating? He later clarified it, and you’re right, he admitted it wasn’t quite what it seemed:
  • The Walk-Back: In a February 12, 2020, New York Post piece, Schnatter said, “I didn’t say I had eaten 40 pizzas in 30 days. I said I had 40 pizzas in 30 days. When I said I had a pizza, it means I’m inspecting. I’m not eating every pizza. I may be eating parts of pizzas.” He framed it as quality control, not a gorging spree—sampling slices, not scarfing whole pies.
  • Plausibility Check: Even “inspecting” 40 pizzas in 30 days is a stretch—over one a day, every day, post-ouster, when he had no official role? He’d sold millions in stock by November 2019, so he wasn’t exactly on the payroll taste-testing. It’s believable he tried some pizzas to fuel his narrative—maybe 10 or 15—but 40 feels like hyperbole to hammer his “it’s not the same” point. No receipts or pizza boxes ever surfaced to back it up.
  • Public Reaction: The clarification didn’t kill the meme. X users like Q_Review (February 13, 2020) quipped, “Imagine a man so filled with spite he orders 2 pizzas every three days to pick apart with his fingers and ‘inspect’—that’s worse!” The damage was done—40 pizzas stuck as a symbol of his unhinged vibe.
So, yes, he exaggerated—or at least overstated—the scope. “Inspecting” parts of 40 pizzas is less absurd than eating them whole, but it’s still a wild flex, likely inflated for effect. He admitted the miscommunication, but the original claim’s absurdity is what cemented it in lore.


Why It Endures
The interview hit in late 2019, pre-COVID, when the world could still laugh at a sweaty pizza guy ranting about betrayal and crust quality. Then the pandemic gave it a dark twist—memes tied his “reckoning” to global upheaval, a cosmic punchline. By March 8, 2025, it’s a time capsule: Schnatter’s reckoning never came, his pizza tally was half-baked, but the internet immortalized both. No legal win, no Papa John’s downfall—just a viral echo of a man scorned, and a lot of imagined pepperoni.
 
If you arent using chatgpt for a replacement for googling/searching forums on how to do well documented things youre really behind the curve.
Time for Null to run our very own, @KiwiGPT. Something that we can tag and can give us a summary of the OP, the general consensus of a thread based on replies, etc etc.
 
Ive started running job applications through gpt, solely to fix commas as our system is fucking convoluted. Usually it helps the flow a little too as I tend to write quirky senteces, but in this case that's good cause itll leave a bit of personality in its corrections.

Honestly I havent even considered putting in an entire cover letter and asking for advice. Hm, I should try that.
 
This is a good use of it:
1743347350052.png
 
ChatGPT is less cucked and retarded than it used to be, but that's a very low bar.

Not really. The output quality is essentially random and based on whatever tokens the ai decided would work best with the tokens in its context based on its settings and training.

Every single one of these uses relying on factual information is retarded. LLMs are more like a fancy autocorrect than some kind of star trek computer. They just regurgitate groups of letters and have no understanding of anything beyond 'these letters go with these letters.' It has absolutely no way of distinguishing between false or true information. The concept of information doesn't even exist for it. Concepts don't exist for it at all. It's just letters in, letters out.
This is correct.


They use reinforcement learning and a whole bunch of stuff. They do not all function exactly the same.

Like they're definitely not perfect but your explanation of them doesn't match with current reality.
This is also correct.

The reason for this apparent paradox is twofold: emergent complexity, and our collective push for scale maximization. I could probably be convinced to discuss it more in the relevant thread, but I assume it's already been brought up there and I'm incredibly lazy. It's quite interesting, though, and gets epistemological pretty quickly.

On-topic for this thread: ChatGPT's very useful as a DM tool for tabletop RPGs. Don't let it write your campaign for you or you're a gay retard, but it's good at generating loot tables and stat blocks for well-documented systems like 40k, 5e, and even some of the more obscure FFG- or GURPS-based systems. Your party decides to devote an entire session to exploring the spooky house you only mentioned to add flavor to the setting? ChatGPT. It's also good for summarizing disjointed notes at the end of a session. Sometimes it's better at understanding what I meant than I am.
 
I sometimes go for Copilot when I want to word a search in a very specific way. I always ask for sources.
I don't know if I'm just getting too used to it but I swear search engines are fucking garbage across the board now. I accept the possibility that my duck-fu is just diminishing though.
It's also good for quick text formatting that I don't wanna bother with myself - "here's this info made up of x, y and z, change it up such according to template n below"
 
AI is great for learning new materials. It's like having a teacher in your pocket you can ask any question, and get a coherent answer. That answer can then lead to you asking new questions, and therefore learning in a personalized, meaningful way.
Pretty much this. Its been a boon for self-study to be able to ask it a detailed question with example and get it to give me something that helps my understanding.

Been using the Brave AI for that, helps with my certificate understanding to say "explain why X and Y are different" or "I understand X but what if its applied in this manner what happens" or just taking practice exams and I'm not sure why one answer is correct over another to ask it to break down the explanation.
 
  • Like
Reactions: Farsided
I've found it's pretty good for tech support. When I have an issue with a system that I'm stumped on, it often finds a decent solution that doesn't take too long to do. A lot of the answers are probably the same as what you could find on forums, but it has saved me hours of searching. You can also upload a huge several megabyte size system log and it will find the solution rather than sift through it yourself and it will find things you probably wouldn't think to look for.

I have an old laptop which has arch on it. Haven't updated in a while (two years) and yesterday when through with updating it. Needless to say, borked it somehow. GPT was very helpful with troubleshooting, way more than those fags from arch.wiki.

It actually helped me understand the errors and pinpoint what was going wrong. Not to mention i used GPT to fine comb stuff before actually upgrading, sorting out conflicts etc.
 
AI is great for learning new materials. It's like having a teacher in your pocket you can ask any question, and get a coherent answer. That answer can then lead to you asking new questions, and therefore learning in a personalized, meaningful way.

Yes, there is the risk of hallucinations, but I actually trust a heartless, unfeeling computer more than I trust your average redditor, which is where you would end up anyways trying to learn from another person. (Or worse, stack exchange niggers, yuck).
I use DuckDuckGo's AI bot for this, since it's free and offers multiple models. General overview of the topic or a starting point by AI, reviewing the professional jargon it uses that I may not be familiar with, then diving into the citations it offers, then following up with a solid reference or a well-reviewed book on the subject.
 
Like what @Oh! said, I've had luck using it for GM resources. It does general set dressing really well. Things that are unimportant, but add depth with a little work.

I'm running a cyberpunk game where the players are currently investigating a gang owned nightclub. Working with ChatGPT to write documentation felt a lot more effective than outright telling players the key details.
The players were handed a physical document that the club's employees have to sign, and they got to pore through it and spot all the predatory clauses that are hiding in the legalese, essentially that employees are signing away their rights, that they'll be 'relocated' to a 'third party' if they try to quit, and that reading the document equals consent, among a few other bullet points I put in the prompt.

It would take me hours to handwrite it and the quality wouldnt even be comparable. They didnt need the whole document, but everyone likes feelies.
 
  • Like
Reactions: Oh!
Back