I am quite sure that in a world where friendly tool AIs were provably easy to build and everyone was gonna build them instead of something else and the idea even made sense, basically a world where we know we don’t need to be concerned about x-risk, Yudkowsky would be far less “relaxed” about AI+power. In absolute terms maybe he’s just as concerned as everyone else about AI+power, but that concern is swamped by an even larger concern.
SarahNibs
Karma: 2,917
What convinced you that adversarial games between friends are more likely a priori? In my experience the vast majority of interactions between friends are cooperative, attempts at mutual benefit, etc. If a friend needs help, you do not say “how can I extract the most value from this”, you say “let me help”*. Which I guess is what convinced me. And is also why I wrote “Maybe I’m bubbled though?” Is it really the case for you that you look upon people you think of as friends and say “ah, observe all the adversarial games”?
*Sure, over time, maybe you notice that you’re helping more than being helped, and you can evaluate your friendship and decide what you value and set boundaries and things, but the thing going through your head at the time is not “am I gaining more social capital from this than the amount of whatever I lose from helping as opposed to what, otherwise, I would most want to do”. Well, my head.
No, that is a cooperative game that both participants are playing poorly.
I believe the common case of mutual “where do you want to go?” is motivated by not wanting to feel like you’re imposing, not some kind of adversarial game.
Maybe I’m bubbled though?
Efficiency trades off with robustness.
If you, the listener/reader, fully understood what I tried to say, it is very very likely that you (specifically you) could have fully understood had I compressed my communication in some ways tailored to you.
collaborative truth-seeking doesn’t exist. The people claiming to be collaborative truth-seekers are lying
Certainly if I wanted to do some collaborative truth-seeking I would choose a partner who believed collaborative truth-seeking existed.
If I didn’t think the possibility for collaborative truth-seeking with a particular individual existed, I would be very tempted to instead just sling gotchas at them.
I tried code interpreter on some of the D&D.Sci challenges here. As expected, it failed miserably at generating any useful insights. It also had some egregious logic errors. I didn’t, but should have, expected this.
For example on https://www.lesswrong.com/posts/2uNeYiXMs4aQ2hfx9/d-and-d-sci-5e-return-of-the-league-of-defenders the dataset is three columns of green team comp, three of blue team comp, and a win/loss result. To get an idea of which picks win against the known opponent team, it grabbed all games with that team participating, found the games where the other team won, and did some stats on the other team’s comp. Except no, instead, it forgot that it had grabbed games where green was that comp and where blue was that comp, so actually it checked for when blue won and did stats on all of those, aka half the “winning opponent teams” were just the original comp. Its analysis included “maybe just mirror them, seems to work quite well”.
https://blog.mrmeyer.com/2015/if-math-is-the-aspirin-then-how-do-you-create-the-headache/
Here is the most satisfying question I’ve asked about great lessons in the last year. It has led to some bonkers experiences with students and I want more.
“If [x] is aspirin, then how do I create the headache?”
I’d like you to think of yourself for a moment not as a teacher or as an explainer or a caregiver though you are doubtlessly all of those things. Think of yourself as someone who sells aspirin. And realize that the best customer for your aspirin is someone who is in pain. Not a lot of pain. Not a migraine. Just a little.
Piaget called that pain “disequilibrium.” Neo-Piagetians call it “cognitive conflict.” Guershon Harel calls it “intellectual need.” I’m calling it a headache. I’m obviously not originating this idea but I’d like to advance it some more.
One of the worst things you can do is force people who don’t feel pain to take your aspirin. They may oblige you if you have some particular kind of authority in their lives but that aspirin will feel pointless. It’ll undermine their respect for medicine in general.
This story was co-written with GPT-4
Halfway through the first paragraph, I said, out loud, “this was written by ChatGPT”. Do you know which bits of the first paragraph were by you vs auto-generated?
Also an extremely important lesson to learn is that toy problems are actually useful, it’s actually useful to try to solve them, their design is sometimes difficult, a well designed toy problem often works better than it seems from a surface reading, and that continually trying to “subvert the rules” and find “out of the box solutions” does not end up getting you the value that the toy problem designer was aiming to give you.
Thinking and coming to good ideas is one thing.
Communicating a good idea is another thing.
Communicating how you came to an idea you think is good is a third thing.
All three are great, none of them are lying, and skipping the “communicating a good idea” one in hopes that you’ll get it for free when you communicate how you came to the idea is worse (but easier!) than also, separately, figuring out how to communicate the good idea.
(Here “communicate” refers to whatever gets the idea from your head into someone else’s, and, for instance, someone beginning to read a transcript of your roundabout thought patterns, bouncing off, and never having the idea cohere in their own heads counts as a failure to communicate.)
[link here once it’s published]
https://www.lesswrong.com/s/zLib3j2Fdnnx3aP3F/p/7oAENKMsud2qQBXDj
FWIW, when I read
You can even do partial runs, e.g. roll the ball down the ramp and stop it at the bottom, or throw the ball through the air.
But you only get one full end-to-end run, and anything too close to an end-to-end run is discouraged.
I heard “you can roll the ball down the ramp and stop it at the bottom, but we will discourage it and look at you sideways and you will get less metaphorical points if you do”.
Discord link invites have expired. Edit: looks like https://discord.gg/bloombotai is the permanent invite.
Q6: Spend fifteen minutes home alone
5.6 to 9.5?? Average 7.4??? These numbers seem ludicrously high.
[SUMMARY] = GuySrinivasan, a user of Lesswrong website, is shocked by an AI-generated story about a woman whose head suddenly became a tomato fruit. He advises the author to try to get a prompt which writes a great story given a REASONABLE premise instead
===
Chapter 1: Synthesis
In the cool, sterile light of his home office, Guy Srinivasan’s eyes skimmed over lines of code on his dual monitors. He was a software engineer by trade, a regular contributor to the Lesswrong website by passion, and an amateur botanist by curiosity. His world was a delicate balance between the rigid logic of artificial intelligence and the unpredictable beauty of organic life.
The house around him was silent save for the soft hum of his computers and the distant rhythmic pulse of the hydroponics system in the adjoining room. The walls, a tasteful silver-grey, were adorned with framed infographics of plant biology and neural networks, an odd juxtaposition that Guy found oddly comforting. From the large window beside him, he could see the cerulean glow of the city’s solar grid, flickering like a neural network against the backdrop of the night.
His fingers moved deftly across the keyboard, crafting intricate algorithms, the language of his vocation. The screen reflected on his spectacles, the code lines dancing like cryptic symbols in his eyes. He paused, leaning back to consider his work, a new AI model designed to generate narratives based on user prompts.
A soft ping echoed through the room, a message notification from the Lesswrong website. A user, TerraNova, had posted a new story generated by an AI model. The premise was absurd—a woman whose head turned into a tomato. He shook his head, a smile tugging at the corners of his mouth. He remembered a time when the AI’s prompts had been… less adventurous.
Tapping a quick reply, he offered a suggestion for a more reasonable prompt. But as his fingers moved, his gaze drifted towards the hydroponics room, where an array of tomato plants swayed gently, their LED lights painting them in hues of indigo and magenta.
The thought lingered as he returned to his work. His model, he mused, was designed to find the harmony between the bizarre and the reasonable, to generate stories that were both imaginative and grounded. But what if the true challenge lay not in the premise but in the execution?
His fingers stilled, and for a moment, his eyes lingered on the lines of code before him. He looked back at the tomato plants, their leaves rustling softly in the artificial breeze. The absurdity of the premise, he realized, was not an obstacle, but an invitation to explore the boundaries of reality in a world bound by code and chlorophyll.
His day concluded with an appreciative sigh, the soft clacking of the keyboard echoing through the room. He saved his work, the lines of code, still incomplete, a testament to the story yet untold. As he shut down his computer, he couldn’t help but wonder. What would his model make of a woman whose head turned into a tomato? He smiled to himself, the question lingering as an echo of the day’s labor and a promise of tomorrow’s exploration.
And so, Guy Srinivasan, software engineer, Lesswrong contributor, and botanist enthusiast, closed his eyes, a single question dancing in the labyrinth of his thoughts, the seed of a story yet untold in the quiet night.
[SUMMARY] = a woman’s head suddenly became a tomato fruit
===
Chapter 1: A Regular Day at the Biomechanics Lab
The sun had barely risen, its light filtering through the thin aluminum blinds of Dr. Seraphina Moreno’s apartment. A mechanical whirring filled the air, her coffee machine autonomously preparing her morning dose of caffeine. She loved the smell, rich and earthy, permeating through her tiny kitchen while she hastily pulled on her lab coat.
Seraphina worked in the cutting-edge SynthBio Labs, nestled within the heart of Seattle’s bustling technopolis. Her days were a heady mix of biomedical engineering and advanced genomics, creating hybrid organisms that had the potential to revolutionize medicine, agriculture, and more. But for her, it was a world teetering on the edge of the miraculous and the terrifying, where every discovery held the potential for unimaginable benefits or catastrophic consequences.
At SynthBio, Seraphina was responsible for the morphogenesis protocols, fine-tuning the genetic blueprint of each organism to give it the desired form. Today, she was working on a tomato plant, attempting to increase its size without compromising its inherent structure.
In her lab, the faint hum of the genetic sequencers and the sterile smell of disinfectant welcomed her. She settled into her work, the soft glow of the holographic display illuminating her face as she navigated the complex genomic map of the tomato.
As she worked, her colleague, Dr. Chen, peered over her shoulder. “Moreno, are you still trying to supersize that tomato?” His voice was tinged with skepticism. The challenges of scale in bioengineering were well-known, and Seraphina’s project had been deemed a fool’s errand by many.
She smiled, the corners of her eyes crinkling. “The potential benefits are worth it, Chen. Imagine the impact on world hunger if we succeed.”
He grunted noncommittally, returning to his own work. They operated in different spheres of expertise: Seraphina, the artist of form and structure, and Chen, the master of functionality and adaptation. It was a dynamic that had sparked many a spirited discussion over the years.
Hours passed in a flurry of data analysis, gene manipulation, and computational modeling. Yet, amidst the calculations and hypotheses, there was an underlying reality she couldn’t escape—the tomato project was proving more complex than she had anticipated. A subtle tweak in one area would trigger a cascade of unforeseen changes in another. Despite her expertise, the project was on the brink of becoming unmanageable.
By the time Seraphina looked up from her work, the lab was dim, the only illumination coming from the soft, pulsating light of the sequencers. Chen had left long ago, leaving her alone with her thoughts and her incomplete genetic model. She stared at the holographic tomato suspended before her, its complex network of genes and proteins glowing with an ethereal light.
Sighing, she saved her work and shut down the system. As she shrugged off her lab coat and prepared to leave, she glanced one last time at her project, her gaze lingering on the floating image of the tomato. It was a challenge, no doubt, but one she was determined to overcome. Tomorrow, she would return, ready to tackle the intricate puzzle of the tomato’s morphogenesis once again.
As she stepped out of the lab, the city’s neon lights danced in the darkness, the technopolis alive and pulsating. The challenges of her work lay heavy on her mind, but she couldn’t help but feel a sense of exhilaration. After all, she was at the forefront of a new scientific frontier, where the lines between biology and technology blurred.
She was shaping the future—one gene at a time.
[SUMMARY] = after the invention of extremely resilient construction materials, humanity has spread to the ocean floor. We are now facing global food crises as the oceans ecosystems are disrupted beyond apparent recovery.
===
Chapter 1: The Depths Below
Dr. Aiden Gallagher pressed his forehead against the cold glass of his submersible, eyes darting over the illuminated expanse of the ocean floor. The abyssal plains of the Atlantic spread out before him, an alien landscape encased in perpetual darkness, punctuated by the occasional glow of a bio-luminescent creature or the distant twinkle of a neighboring underwater city. His city, New Atlantis, hung suspended in the depths, a lattice of gleaming domes and dark spires that defied the crushing pressure of the ocean.
He exhaled slowly, watching as the moisture from his breath fogged up the glass. The submersible was silent except for the hum of the engine and the gentle hiss of the oxygen recycler. A lifetime of marine biology hadn’t prepared him for the desolation that met his gaze now. It was the weight of the world, pressing in from all sides.
He was shaken from his thoughts by the crackle of his comms unit. “Control to Gallagher,” the voice said. “You’re drifting, doctor. Please confirm status.”
“I’m here,” Aiden replied, adjusting his course. His work was crucial in understanding the ecological shifts occurring in the ocean due to humanity’s migration underwater. The food web was a delicate tapestry, and it was beginning to unravel at the edges.
Hours later, he docked his submersible at the New Atlantis research station. The large, dome-shaped facility was a hive of activity, people in lab coats scurrying about, screens blinking with data, the low murmur of serious conversation filling the air. Aiden’s home, his world.
He was met by Dr. Rina Patel, his long-time colleague, and friend. Her dark eyes were serious, lined with worry. “Another barren sector?” she asked, no preamble.
Aiden simply nodded. The implications were clear.
That evening, Aiden sat at his desk, poring over the day’s findings. The silence of his office was a stark contrast to the busy station outside. His fingers traced the topography of the sea floor on his holographic display, the barren areas growing with each passing day. It wasn’t just about the fish, the coral, or the microscopic phytoplankton – it was about humanity. If the oceans died, so would they. The stakes were as high as they could get.
He stayed there for a long time, staring at the hologram, lost in thought. Then he leaned back in his chair, rubbing his tired eyes. The enormity of the problem was overwhelming, yet he could not afford to look away. This was his world, his responsibility. He made a silent promise to the cold, dark ocean depths outside his window. He would find a way. He would fix this.
Finally, he shut down his display and left his office, the station’s lights dimming as the artificial night cycle began. As he walked the silent corridors, his mind teemed with ideas and unanswered questions. A single day had passed in New Atlantis, but for Dr. Aiden Gallagher, it felt like a lifetime. He knew he was standing on the precipice of something immense. The ocean had always been a mystery, a siren’s song luring him into the depths. Now, it was a challenge. And he was ready to dive headfirst into the unknown.
In this top-level comment I will give my prompt. In each of three comments replying to this one I will give the one-shot results of the prompt on three premises. If I update this comment with a new prompt I will also update the replies with new results. Feel free to crib. I began with the OP’s sample prompt, used a reasonable premise for the story, revised the prompt to avoid the most egregious writing and stylistic mistakes, and inserted a bit in the last paragraph I stole from Christopher King.
PROMPT (replace [SUMMARY] with a summary of the novel):
You are writing a hard science fiction novel. A summary of the entire novel is that [SUMMARY]. Amidst this backdrop our characters’ stories are spun.
In the opening chapter of your hard science fiction novel, set the stage by actively showcasing the protagonist’s world and their role within it. Immerse the protagonist’s expertise and experiences within the narrative and subtly imply their part in a larger system through their interactions and tasks. Avoid urgency; the story is only beginning and the exposition is not deep enough to support organic urgency in the first chapter.
Give your setting a vivid description, allowing the reader to envision the unique aspects of your world and the challenges it faces. Convey these challenges through the actions, dialogues, and experiences of the characters, including how different stakeholders handle the situation.
Touch on the events going on in the protagonist’s life which make this character the lens through which we will explore the world and plot. If an issue connected to the plot is shown through the chapter’s events, don’t reveal its full implications. Let the readers mull over its meaning as they progress through the story.
Conclude the chapter without any material change in tone. The protagonist should finish something, like their day, or a report, or a discussion with another character. The reader should leave the chapter engaged and curious, wondering about the science fictional vista just explored.
Ensure that every scene contributes to the overall narrative and character development. Maintain a pace suitable for a longer work, allowing the plot to unfold gradually. Avoid divulging all the plot or world details in this initial chapter; instead, plant seeds of information that will germinate later in the story.
Foster an environment of intellectual challenge and curiosity. Engage the reader with complex characters, nuanced humor and wit, and profound, unexpected ideas. Your story should delve into its themes subtly, presenting a world grappling with significant challenges without resorting to melodrama or over-emotion.
Remember, you’re writing hard science fiction. Equally important as character development is the detailed world-building and intricate plot. Every scene, whether high-stakes or quiet, should contribute to the overall narrative and character development. Aim to create an original, creative, and captivating story that keeps readers intellectually stimulated.
Writing Advice:
Maintain subtlety and restraint throughout your narrative. Communicate the story’s stakes implicitly rather than explicitly, and avoid melodrama or overstatements that could disrupt the tone or force the plot.
Depict the protagonist’s reactions or emotions by showing their internal processes without resorting to overt dramatization. Let the reader infer the situation’s magnitude through the character’s actions, decisions, and behavior changes.
Trust your reader as an active participant in your story. They are intelligent enough to pick up on clues, make connections, and recognize the events’ significance without explicit cues.
Ensure the stakes are grounded in your world’s reality and the experiences of your characters. Let the situation’s severity be understood through its impact on the world and the characters’ lives, and through the complexity of the challenges they face.
In hard science fiction, realism and believability are key. The reader should feel the tension and urgency because they understand the intricacies of the situation and empathize with the characters’ struggles. Drama should arise naturally from the plot and character dynamics, not from overt expressions.
More Writing Advice:
Let your narrative breathe. Don’t rush to escalate the stakes or dramatize the events. Allow the tension to build gradually and organically.
Hint at the situation’s severity subtly through your characters’ reactions, thoughts, and decisions.
Trust your reader’s intelligence and their ability to understand the story’s implications and complexities.
Keep the stakes grounded and relatable. The reader should feel the urgency and tension because they understand and empathize with the characters’ struggles.
Conclude the chapter gently, grounded in current events, from a third person limited perspective.
In hard science fiction, the extraordinary circumstances should be rooted in the reality of your world and the lives of your characters.Choose a distinct writing style and stick with it. Humans reading the chapter should not suspect that it was written by a LLM. Only write the novel excerpt, do not describe what was written or tie to back to the writing advice or these constraints.
One of the most valuable things I have done, for myself, is to let as much of my personal life bleed into my work behaviors as I can, as you define them.
This could have backfired spectacularly. In some work cultures probably it would always backfire.
In mine, I:
make 98%+ of my writing viewable to everyone at the company, and we’re remote, so almost everything of importance makes it into writing
never “try” to display an air of competency—trying to display an air of competency is one of the core behaviors that caused terrible feedback loops and major depression early in my career, now I take joy every time I can display to everyone where I am not competent. In some sense this is signaling extreme competency because who would do that unless they were very comfortable in their position. See also “backfire”. But also this can lead to much more rapid professional competency growth, because other people love to teach you things.
tell jokes, embarrass myself a little, feel okay being silly or weird, literally treat it as a red flag about a person if I feel I need to walk on eggshells around them and bring it up with my manager even if I can’t point to exactly why
push for exploratory “something seems interesting here but IDK what and no I can’t tell you its value” work in general, and in specific do some of it myself whenever the mood strikes and nothing urgent is otherwise going on