Jump to content

Captainant

Certifiably Surly
  • Posts

    18149
  • Joined

  • Days Won

    6

Everything posted by Captainant

  1. BattleTech! My new LGS has a good group running the Battle of Tukayyid as part of a big ass Clan Invasion campaign. I'm pretty familiar with the breadth of the rules from playing for decades with my dad and buddies, so I think it'll be a good acid test for something that's actually useful. My M4 MBP should have enough juice to run it all locally - at least that's the theory I'm testing You can spend a BUNCH of time trying to find the right rule for this specific instance, and it would be nice to have a citation machine that can point out the specific rules that tell me what modifiers I get, or if there REALLY is a rule for ripping your opponents mech arm off and beating them with it
  2. If you can lingustically frame it such that you have it acting as a "GM" for a tabletop game, or frame it as a fantasy setting that removes the real world context, the models guardrails get fuzzy extremely fast. If you're asking those questions only to the hosted chat application, you're going to have a different experience from directly querying the LLM that underlies the chat apps of OpenAI, DeepSeek, Ollama, et al. The hosted chat applications basically all have a postprocessing layer to check for the case you're describing, but the models that large enterprises host don't. That lack of built-in guardrails for the technology is a big part of why businesses that are chasing """AI""" workloads are deferring to foundational model providers rather than assume that liability themselves. Granted, the infamous $1 car sale happend on openAI's chatGPT application, but that was prior to many of the guardrails that now exist as standard
  3. when you start to use a metric as a goal or target, it ceases to be a useful metric precisely because it starts to be manipulated Again - chatGPT is NOT a large langue model. It's an application build on top of an LLM. If you are running an LLM yourself, you can give it whatever direction you want and override any sort of moral framework that the authors tried to instill into it. In this case, chatGPT has those guardrails hardcoded into its prompts and also sitting between the LLM response and the user, in case someone "jailbreaks" the model.
  4. This is a garbage in::garbage out problem. The LLM's have that tendency because it existed in the training data used to build its probabilistic model of natural language interactions. It feels like it's just scaling up the existing problem with outsourced technical solutions. Yeah, you might get something that works, but it's extremely narrow to just that context. If you try to get it to build something general for multiple problems, you find yourself spending as much time as just developing it yourself. Except you're spending a shitton more in resources to get the same result
  5. HTTP 411 error
  6. it seems like Twitter/X and Insta have been having some serious embedding issues lately, and have been adopting dark patterns like redirect loops to make it harder to leave their site
  7. I've gotta steal an excellently cited post that lays out the timeline with contemporaneous news or court filings cited for each timeline event: February 16, 2017: Alex Acosta nominated Secretary of Labor. July 2, 2019: Jeffrey Epstein indicted. July 12, 2019: Alex Acosta resigns. August 10, 2019: Epstein dies by suicide. June 20, 2020: Trump fires Geoffrey Berman from SDNY who indicted Epstein June 29, 2020: Ghislaine Maxwell indicted. March 29, 2021: Maxwell's Superseding indictment. November 16, 2021: Maxwell Jury selection begins. December 29, 2021: Maxwell convicted on 5 of 6 counts. February 28, 2023: Maxwell appeals. September 17, 2024: Second Circuit rejects appeal. January 15, 2025: Maxwell delays SCOTUS appeal until Trump is inaugurated March 2025: Pam Bondi conducts emergency review of Epstein and Maxwell documents. April 10, 2025: Maxwell files cert petition at SCOTUS May 7, 2025: Trump's Solicitor General John Sauer delays SCOTUS response. June 6, 2025: Trump's Solicitor General John Sauer delays SCOTUS response again July 7, 2025: Pam Bondi claims there’s no there there. July 8, 2025: Trump loses it over questions about Epstein. July 12, 2025: Trump attempts to claim Epstein is a Democratic plot. July 14, 2025: DOJ defends Maxwell prosecution at SCOTUS July 15, 2025: WSJ interviews Trump about Epstein Birthday Book July 16, 2025: Pam Bondi fires Maurene Comey (Maxwell's prosecutor), on Trump’s personal authority. July 17, 2025: Trump yells at supporters who won’t move on from Epstein. July 18, 2025: Trumps defense lawyer Todd Blanche files to unseal Maxwells grand jury materials July 22, 2025: Mike Johnson shuts US House until September to avoid Epstein vote
  8. Wow, Congress just shut down until September to avoid voting on the record on releasing the Epstein files. Totally nothing there though, let's get on to the important business like making the commanders redskins again
  9. it's incredibly frustrating that they're slashing NASA's budget but allocating money for boondoggles like this that do NOTHING to further their science mission
  10. I'm more annoyed about the frequent flier new account because I can't fucking block you. You just come back again and again and again despite there being a rule specifically against YOU for your disinformation antics during COVID and the early stages of the russian invasion of Ukraine.
  11. So now that we're at the point of ICE entering schools and their coaches standing up against the feds saying they'll die before they let their kids be taken, I hope you're paying attention.
  12. Some of my favorite childhood family vacations were doing bike rides through Glenwood Springs to Hanging Lake and Spouting Rock, and then going on to visit my granddad in Crested Butte. Man, that's such a beautiful place! If you can swing it when the kids are older, you should try and make a stop at the Great Sand Dunes to take your kids sledding. It blew my mind to go sledding in July lol
  13. love it when fascists face consequences
  14. That's specifically NOT a large language model though - it's an agentic design where an LLM takes your ask for information, decomposes the query into smaller sub-queries, and then searches the internet on your behalf and creates a result based off those findings. It's a just-in-time retrieval augmented generation solution, with the search agent powered by the Gemini LLM foundational model. That specific solution is literally just a tech demo I show to customers who need an idea of what LLM's can do. It turns out it's incredibly powerful to do feature extraction on your corporate corpus of documents to have a corporate rule citation machine - I'm also working on building a local RAG engine (I'm not a genius, I'm just following a github repo) for tabletop game rules that I'm super familiar with as a "this can't work, can it...?" In google's case their big value add is that they already own a MASSIVE index of information and documents from crawling the web for literal decades. In both cases, the LLM is the mechanism that's being used to sift through the quantity of information that is otherwise too dense for a human to parse. THAT is the core utility of LLM's, and something that I don't think is ever going to go away. It's just that's NOT what is being sold right now. There's alot of pixie dust and snake oil being peddeled.
  15. Nevermind that the administration is now fully pantsed on their lies to the court previously insofar as "not having jurisdiction there"
  16. I've been listening to the interview off and on, and he's really on the money about the disinformation storm. you should watch the interview (or wait for clips to come out, whatever lol), considering he's been trying to go the fuck away and yet he's still actively a target of MAGA. The fact that you feel that way reflexively is a strong indicator of how successful their messaging has been bingo, the entire point is that the fascists need SOMETHING to distract the idiots and get their amgydala pumping. Nevermind that trumpco is doing EXACTLY what they've been investigating Hunter Biden for over bullshit and baseless allegations.
  17. ICE - coming to a little league field near you!
  18. I genuinely am interested to hear your experience using bitcoin for commerce, and your general experience using it in contrast to traditional financial systems
  19. My comment on your water value was about as serious as your initial comment was lol
  20. You're betting the don't pass line, and cheering on the seven out. Dahobbs pretty succinctly said my feelings on its inherent unsustainability. Also lmfao what is this, your 40th new account? In a sense, investing in Bitcoin to further goose the electricity and water consumption of the network is enhancing the value of your water holdings by inducing further scarcity. Good stuff.
  21. It's also what happened in 1930s Germany. Lots of extralegal action that was oopsie poopsie cleaned up after causing extreme hardship to the untermensch classes. Their media also swept it under the rug, for fear of reprisals from their populist authoritarian leader. That was the nice phase, before it became open warfare
  22. https://www.theguardian.com/us-news/2025/jul/21/migrants-miami-ice-jail-abuses https://www.hrw.org/report/2025/07/21/you-feel-like-your-life-is-over/abusive-practices-at-three-florida-immigration Migrants at a Miami immigration jail were shackled with their hands tied behind their backs and made to kneel to eat food from styrofoam plates “like dogs”, according to a report published on Monday into conditions at three overcrowded south Florida facilities. The incident at the downtown federal detention center is one of a succession of alleged abuses at Immigration and Customs Enforcement Agency (Ice) operated jails in the state since January, chronicled by advocacy groups Human Rights Watch, Americans for Immigrant Justice, and Sanctuary of the South from interviews with detainees. Dozens of men had been packed into a holding cell for hours, the report said, and denied lunch until about 7pm. They remained shackled with the food on chairs in front of them. “We had to eat like animals,” one detainee named Pedro said.
×
×
  • Create New...