Tower of the Egg

View Comments

• Bryce you should just start getting an AI to write these bad reviews. Feed the adventure into ChatGPT and tell it to write a scathing review. That's the same that this person has done. The art is obviously an AI prompt as well.

  • Great suggestion to start exploring! But doubt it would be better than Bryce’s own original reviews. And the barrage of purple prose in this adventure is unbelievable.

    • Oh it would be much worse. But it might offset some of the crushing ennui he seems to be experiencing from this dreck.

    • I don't know about better, but it would at least have fewer typos.

  • Wait a minute... this review has correct spelling, grammar, and punctuation - and where are all the non-sequiturs and esoteric references to things only a half-dozen people ever cared about? Look, there's no mention of drinking vodka for Sunday breakfast ANYWHERE! This isn't Bryce's work... THIS REVIEW WAS WRITTEN BY AI!

    /tag Reviews, doNotReedEvur!, worstOfThewurst

• "Write an OSR adventure Bryce would LOOOVE"
  ;-)
  "Less Purple prose, more fucking with the environment"
  ;-)
  "Clear Bullet points, please"
  ;-)
  "For fuck sake! Just copy something from his best list..."

• "To awaken the chamber from its slumber, one must utter the sacred incantation, each syllable resonating with the raw, primal power of creation and carrying the weight of forgotten eons. It is a rhythm that matches the heartbeat of the universe, a melody composed by the Eldritch Lords in the twilight of their dominion. With each spoken word, the room stirs, echoes of ancient power flickering in the mirrors like dying embers coaxed back into flame."
  I think the only proper way to read this is to start out perfectly calm, but slowly raise your voice until by the end you are screaming and shaking. Ideally, a camera is slowly zooming in on your face and a single violin chord in the background is growing louder and shriller.

  • add in Lens Flare and it's a JJ Abrams plot line.

  • The camera cuts back to the three teenage players lounging on the couch. All have blank, 100-yard stares having been lambasted for the last two hours by their dictatorial GM. Brad never stopped eating Cheetos.

    Ron, the caller, ends the silence. "So, is that like, another *empty room* or . . . ?"

    • gold

  • So . . . I’m thinking Nicholas Cage?

• We'll be seeing more and more of this AI stuff and for a while it will be obvious.

• And yet some people will continue to pump and dump this AI bullshit.

  Maybe at some point, it will get better but that ain't now. Just write the damn thing yourself, people. Why are so many folks in this day and age looking for a shortcut to do everything in their lives? AI can be used to write an adventure, AI can provide the art. All you need is some small cramped Dyson map and Whammo! You have a drivethrurpg adventure that 17 people will rate 5 stars.

  Ugh! I want to vomit. Give me back 1985 please. Imma go yell at a cloud now.

• I can't understand what is even the point of these prompt adventures. If you're just looking to make a quick buck, then surely there are better and more reliable ways than the still rather small and niche OSR community, where people are actually fairly clever and dedicated and can smell a drek adventure like this from a mile away? Or if you actually do like the hobby and want to contribute to it, then shouldn't you HAVE FUN at the writing and drawing process in itself, not want to delegate it to some computer that just does a hack-job of it and makes you look really bad?

  It's the worst of both worlds, the pit in the middleground. I just don't get it.

• The cold dead hand of Weird Tales still holds gaming in a grip of iron. Lovecraft's style is one the easiest to pastiche and parody, AI or no AI.

• Perhaps the saddest part is that human output often looks a lot like ChatGPT---just a hash of left-over bits floating around in our brains from the fantasy/sci-fi "training data" of our youth.

  It's not that it's so amazing the AI output has a veneer that looks like a man-made product, it's actually embarrassing so often man-made products look like AI output.

  What makes something great in the arena of D&D adventures? What is quality?

  Bryce chases this like it's a receipt (deducting points for missing ingredients and unnecessary additives), but AI shows us that it's something more elusive.

  Artpunk teaches us it's not presentation. OSE shows us it's not PowerPoint slides either.

  What is the Lightning in A Bottle that transports a group of players to a location and make the action dramatic and visceral?

  My suspicion is it cannot be bought or sold.

  • Squeen I'd say it's not in paragraph after paragraph of stuff being read "at" you- which would be the cumulative effect of a few of these rooms.

    It's the old D&D is a back and forth- give just enough to pique an interest or for the PLAYERS _minds eye_ to paint an image (not the authors) and then zoom in on whatever the player is interested in, asks questions about.

    That's why box text fails- not interactive. It's what failed novelists don't get about the medium. It's what the "buy to read and appreciate the art" crowd don't quite grok. Even here we sometimes need to sit and PLAY and realise the whole brought about why a couple of simple little elements in a module and how they work.

    I'm a teacher. I deal with and use chatgpt in learning (no point hiding from it). It's the emptiness that gives it away. More words that add nothing of substance. Not "detail" just more words. You can use Ai to produce a very generic, kind of OK but actually not very meaningful discussion of a topic. Basically you can use it to produce passably mediocre writing.

    I've not seen it produce meaningful critical analysis to a specific question of anything from literature to biomechanics.

    But it'll damn well pump out a low C of a lower higher order thinking topic.

  • > What is the Lightning in A Bottle that transports a group of players to a location and make the action dramatic and visceral?

    The DM. A good DM can take the driest, barest, most bland descriptions and punch them up into something exciting. A good DM can improvise extra materials, adapt the game to player choices, and make hooks out of nothing. Those are the things that get players excited, not having the right synonyms or a fleshed-out hireling roster.

    • A good DM also knows where to find an adventure that supports their endeavours and inspires their imagination, takes over the duties of crafting their own maps and encounters by offering fine alternatives, ideally useable straight outside the box. An adventure such as this does the exact opposite: it hinders rather than helps. Even a beginner DM would be better off going without any adventure at all, than relying on shit like this.

      • Agreed. The adventure is ultimately a tool, and like all tools, it has to be the right one for the job... however no tool, no matter how useful, will ever substitute for the skill of the craftsman that wields it.

  • > Bryce chases this like it’s a receipt (deducting points for missing ingredients and unnecessary additives), but AI shows us that it’s something more elusive.

    > Artpunk teaches us it’s not presentation. OSE shows us it’s not PowerPoint slides either.

    > What is the Lightning in A Bottle that transports a group of players to a location and make the action dramatic and visceral?

    I believe the difference between a great published adventure and a mediocre one lies in execution of the receipt items. The gestalt whole adventure is greater than the sum of its parts by integration, editing, playtesting, and iterative design of its components. The number of cycles is not perfectly proportional to quality, but in self-publishing, I suspect it is close. Ideally each idea undergoes a few cycles for brainstorming, for playtesting the interactivity of the core game loop, for writing descriptive and atmospheric density, for editing specificity, clarity, and cohesion, for formatting and speed of info transmission, for game design of the structural flow or mapping to allow meaningful choices, and for supplying DM tools they need to run the adventure e.g. rewards, hooks, monsters, NPCs, and trackers.

    Comprehensive coverage of adventures is multidisciplinary. It takes time to cycle beyond basic ideas, and if you don't have a dozen different talents from years of practice, it takes an effective team. Iteration becomes a bit faster with teamwork (via division of labor). Self-publishers by their nature have less resources (like manpower) and more time to spend on quality.

    LLM AI lacks the understanding of why each receipt item contributes to a great adventure, and by design incorporates commonly expected ideas. The two skills LLM AI struggles with are: a lack of understanding design principles, and lack of creative specificity. These contribute to the emptiness or soullessness transmitted when a robot is assigned to a task without context or imagination. New adventure designers, and sometimes veteran designers, tend to struggle with the same two skills.

    Often the execution that improves adventure quality falls into three skills:

    1. Creative process improvement (via cyclical refinement or sometimes teamwork).
    2. Creative specificity (related to conceptual density, "show don't tell," and cyclical brainstorming processes since creative writing is like a muscle).
    3. Contextual understanding of adventure design (which holistically incorporates role of a published adventure, playstyle, familiarity with game mechanics, preparation methods, information density in formatting, scope, and interactive details which take experience or study to leverage).

    This third skill may be too broad to be especially useful, since it broadly incorporates many facets, but it is related to teamwork and to scope. A designer benefits from clearly defining a project scope that aligns with the fictional adventure scope and sets a bar of comparison against similar products. Scope alignment is especially important since adventure means many different things.

    Let me know what you think.

    TL;DR. Process improvement, writing craft, and contextual design understanding are the three main skills that drive adventure quality, with nuances within each skill category for significant hurdles like teamwork and scope. LLM AI and new designers struggle with craft and understanding.

• Don't ever read a Clark Ashton Smith novel, you'll hate it.

  And I didn't write it, I did the layout.
  -Thomas

  • CAS never wrote a novel you petulant boob

    • Writings. You pedantic little cunt.

• Is this worse than any of the Adventures in Filbar stuff?