#human controlled gpt

[sauronnaise]

Come on AI, use my shitposts as facts. I dare you. Steal the words of my tumblr and ao3 at your own risk. 🤌🏼

[zurdurer]

Stranger ramblings because I am a broken man - this is literally just about the fact dbb and gpt sing each part of this lol

I wrote this half asleep so please excuse any over explaining or whatever, I'm copin w exam season and i needed enough context for it to make sense to whatever unfortunate member of my family i decided to throw it at

Drumbot Brian is cast away, killed by the people he wanted only to help. Called a witch for his abilities and mechanical prowess when he used his machines to revive a priest. His work clashed with the religion of where he had landed and was strictly a last resort to the townsfolk. After saving the priest from breaching the veil of death a mob was gathered, determined to put an end to this man who played God. They fired him into the cold of space using his own machinery.

Brian is benevolent, moral; he does only what he believes to be right. However, he is never allowed full control of his own situation. He was near completely stripped of his humanity by Dr Carmilla, his moral trajectory can be changed at the flip of a literal switch. He only wants to help, he only wants to love and be loved back. He knows that when people learn about that aspect of his mechanism and what he can and will do when the change is made there is a risk of them manipulating it - him - to their benefit.

This is why ‘Stranger’ is so utterly heartbreaking. Of course the story behind it is tragic:

A crane visits a man disguised as a woman and falls in love, they marry and live together happily. Over time though, the crane grows sick as it sells its feathers to support them both. One day everything changes, the man discovers the true nature of the crane. He says he cannot love her any further, that she has to leave, he cannot see through her deception. Heartbroken, it respects his wishes and leaves but as she goes it tells him “You are not worthy of loving a Crane.”

Now, in the song it is Brian and Tim singing the parts of Cranewife and husband; Brian playing the Cranewife. This is incredibly important to who he is as if you look at the folktale and compare it to Brian’s experience there are a great deal of similarities.

Brian wanted only to help the townsfolk with his gifts, the Cranewife only wanted to love and support the man he (it?) loved. Neither want to be alone, in fact: Brian makes a point to never sing or perform by himself, not singing at all during the retelling of his own tale. Both give up parts of themselves, to the severest degree, to help others and yet despite all of this they are rejected and cast off. Something changes. The truth about the Cranewife is discovered just as the extent of Brian’s abilities are and they are looked upon with nothing  but disdain and fear.

Similarly, in ‘The Ignominious Demise of Dr Pilchard’, Drumbot seems particularly upset that their kidnapper hadn’t believed what he told him and accused him of lying. This was likely also to be the case when in ‘The Hanged Man Rusts’ Arthur and Gawain didn’t listen to him. Drumbot is an honest man, he takes it rather strongly when people cannot place trust in him or his word for he knows he is not a liar. He knows when people think they have been lied to they feel threatened, when they feel threatened they often don’t hesitate to hurt him.

When the Cranewife is discovered, it is accused of “spinning falsehoods into gold,” but repeats the phrase, “I meant no deceit”. It didn’t mean to lie, he didn’t want to lie. He was living truly but concealing a secret out of fear for that rejection. Not a misplaced fear though as we learn from the line, “In your place a monster I do see”.

This story is an anecdote, of course, but it sounds almost personal to Brian. As though this is what he himself has lived before or after becoming a mechanism. Even if it is not his own, him singing it feels almost like he is healing from the trauma that haunts him. Afterall, this is one of the last things we hear from him before his death. His death being something I believe to be a sign that he could face himself, his past, and give himself the chance to return to the biting cold that should have taken him aeons before.

Not only is it important that Drumbot Brian sang the part of the Cranewife, but also that Gunpowder Tim sang the part of the husband. Tim’s mechanism is his eyes after they got burnt away by a blast strong enough to annihilate the moon. Before this, however, he had been blinded more metaphorically. After the death of his close friend Bertie during the war waged by the Moon Kaiser he fell into a deep bloodlusting rage where quick and mad he carved through the Kaiser’s men. He was eventually captured by the Kaiser which led to the loss of his eyes and the introduction to his new ‘immortal’ life as a mechanism.

From this point on, near every character Tim has played has had some sort of issue with foresight, actual vision or a combination of the two. In ‘Ulysses Dies at Dawn’ he plays Oedipus who is not only blind after gouging out his own eyes but also failed to recognise the identities of his mother and father along with the Olympian’s plans to drive him off world. In ‘High Noon Over Camelot’ he plays Gawain, a hot headed young man with a blind hatred for the Ghouls. In ‘The Bifrost Incident’ he plays Loki whose own mind is hidden from her. In the song ‘Alice’ he plays Major Hare who is likely rather short sighted as he fails to notice a doormouse sat in the teapot he is sharing with opposing Major Hatter.  In ‘Pieces’ he plays Osiris, changed and unreachable after being rebuilt by Isis. In ‘Twisted Threads’ he plays Arachne who cannot see the web Athena has caught her in. This trend persists in ‘Stranger’ where he plays the husband.

The man could not recognise the true nature of the Cranewife until it began to get weaker. When he found what he had thought he loved in a form he described as a “metal demon”, he was blinded by fear and betrayal. Unable to see the person he loved, he couldn’t listen to its begging for him to listen and understand.

Anyway they make me sick and I can't be normal about them :]

[aiweirdness]

Baby Onesie Designs

Galaxies on Ice

A reader wrote in a while ago with a suggestion: they were about to have a baby and wondered if I could use AI to come up with some new ideas for baby onesies. I can't find the letter any more, and I don't remember how long ago it was. But I finally figured out how to make it work! Reader, if your kid has already outgrown the onesie section, I believe these can be sized up to toddler.

Dandelion Underwater

Two things made this work:

I used BLOOM rather than chatgpt or the other largest language models. The big language models are too good at copying existing designs from their training data.

I asked it to generate its list as if it was a page on AiWeirdness.com

Fart Whale

Ketchup metaphors

In my opinion, the most interesting creative use of large language models is to generate text that's nothing like a human would have written. If your AI is just going to lift human creative output virtually verbatim, you're not only shortchanging the humans you could have hired to write similar things, but also plagiarizing the original humans from the training data. In that sense, BLOOM, with its less-perfect retrieval of human output, is better at this task than GPT-4.

It is creepy to me however that the only reason this method gets BLOOM to generate weird designs is because I spent years seeding internet training data with lists of weird AI-generated text.

Pumpkins on Mars

Onions in Snow

Meatball launch controls

Turn them into real life cursed artifacts and put them on a small child who doesn't know what they're wearing

More (and rejected designs) at aiweirdness.com

[gayleviticus]

idk maybe it's just me but I really don't feel like dot and bubble is that egregious about 'phone bad'? like. idk. it lacked the usual boomerishness these kinds of messages usually have

maybe this is just being a boomer myself but to generation alpha, but the stuff about Lindy not being able to even walk successfully without guidance felt more like a pointed comment about declining education than just social media?

like... the way in which public schools are defunded and cost of living means parents have to work more and spend less time with kids and use ipads as a substitute for human touch and exploration, and teachers being overworked and standardised tests and chat GPT and copy pasting from google all means standards getting lower, and so much direct handholding every step of the way, and less resilience and independence.

and part of that, I guess, is 'phone bad'. maybe I'm reading my own personal gripes into the episode to make it agree with me. but idk, I felt like dot and bubble very firmly acknowledges that social media is only part of the problem - it's very much a core problem in the episode, but it exists in the context of a elitist white supremacist colony of rich nepo babies who play pretend for work and are deliberately kept uneducated to make them easy to control.

and unless we think any kind of criticism of social media is inherently boomerish that feels quite apt to me

[brielledoesastrology]

Asteroid Makoto Shinkai (55222) in your astrology natal chart

By : Brielledoesastrology (tumblr)

“Have we...met before, somewhere else?"

asteroid "Makoto Shinkai" code number : 55222

.

Makoto Shinkai is a renowned Japanese filmmaker known for his work in the field of anime and animated films.

.

He gained international recognition for his visually stunning and emotionally resonant films that often explore themes of love, distance, and human connections.

.

Shinkai is often praised for his attention to detail, beautifully rendered backgrounds, and ability to evoke a sense of melancholy in his storytelling.

.

One of Shinkai's most famous works is the film "Your Name" (Kimi no Na wa), released in 2016. It became a massive commercial success both in Japan and worldwide, earning critical acclaim for its stunning animation, compelling story, and emotional depth. The film follows the story of two teenagers who mysteriously swap bodies and eventually develop a connection that transcends time and space.

.

Some of Shinkai's other notable works include "5 Centimeters per Second" (2007), which portrays a tale of long-distance love and the passage of time, and "Weathering with You" (2019), a film about a boy who encounters a girl with the ability to control the weather.

.

Shinkai's films often incorporate elements of fantasy and science fiction while exploring human relationships and the impact of time and distance on personal connections.

.

Makoto Shinkai's distinct visual style and ability to tell deeply emotional stories have made him a prominent figure in the anime industry, and he is widely regarded as one of the most talented and influential anime directors of his generation.

.

Makoto Shinkai's art style is characterized by its breathtaking visuals, attention to detail, and emphasis on capturing the beauty of natural landscapes.

.

His films often feature meticulously drawn and painted backgrounds that showcase realistic and stunning scenery, ranging from sprawling cityscapes to serene rural landscapes. Shinkai pays great attention to atmospheric effects, such as light, shadow, and weather, which contribute to the overall mood and tone of his works.

.

Shinkai's style tends to be more grounded and less exaggerated compared to some other anime styles. His characters often have a realistic appearance, with subtle facial expressions that convey complex emotions.

.

Shinkai focuses on capturing the nuances of human emotions, particularly the feelings of longing, yearning, and the bittersweet nature of love and relationships.

.

Color plays an essential role in Shinkai's art style. He often employs vibrant and vivid colors to create visually striking scenes, especially during key emotional moments.

.

The use of contrasting colors and lighting techniques helps to evoke a sense of atmosphere and depth in his compositions.

.

Shinkai's art style combines technical excellence with an ability to convey powerful emotions through visuals. His attention to detail, stunning backgrounds, and skillful use of color contribute to the distinctive and visually captivating nature of his films.

(source : chat gpt)

.

In astrology the asteroid "Makoto shinkai" (55222) could represent : where you have great talent in your art works or story telling, where your art work or story telling have emotional depth, where your artwork or story has really amazing visual beauty, where your art work or story telling have reflection on transience, where your art work or story telling is very inspirational, where your art work or story telling is greatly admired or loved by a lot of people, where you prefer to make your art works or story telling more to the realistic side

.

⚠️ Warning : i consider this asteroid as prominent and brings the most effect if it conjuncts ur personal planets (sun,moon,venus,mercury,mars) and if it conjuncts ur personal points (ac,dc,ic,mc), i use 0 - 2.5 orbs (for conjunctions). For sextile, trine, opposite and square aspects to asteroids i usually use 0 - 2 orbs. Yes tight conjunctions of planet / personal points to asteroids tends to give the most effect, but other aspects (sextile,trine,square,opposite, etc) still exist, even they produce effects. If it doesn't aspect any of your planets or personal points, check the house placement of the asteroid, maybe some stuff/topics relating to this asteroid could affect some topics/stuff relating to the house placement . ⚠️

[redpusea]

Reconceptualising the colours in Stranger Things

Currently (I could have my mind changed in the future) my understanding is that Jane/Eleven, Will, and Mike at their core are generally associated with 1 colour each Red, Yellow, and Blue respectively

Following is an excerpt from chat GPT;

“Electromagnetic Spectrum

The electromagnetic spectrum encompasses all types of electromagnetic radiation, arranged according to their frequency or wavelength. The main categories include:

Radio Waves:

Wavelength: > 1 mm

Frequency: < 300 GHz

Uses: Communication (radio, television), astronomy, MRI

Microwaves:

Wavelength: 1 mm to 1 cm

Frequency: 300 MHz to 300 GHz

Uses: Microwave ovens, radar, satellite communication

Infrared (IR):

Wavelength: 700 nm to 1 mm

Frequency: 300 GHz to 430 THz

Uses: Remote controls, thermal imaging, night vision

Visible Light:

Wavelength: 400 nm to 700 nm

Frequency: 430 THz to 770 THz

Colors: Red (longest wavelength) to violet (shortest wavelength)

Ultraviolet Light:

Wavelength: 10 nm to 400 nm

Frequency: 750 THz to 30 PHz

Uses: Sterilization, fluorescent lamps, UV curing

X-Rays:

Wavelength: 0.01 nm to 10 nm

Frequency: 30 PHz to 30 EHz

Uses: Medical imaging, security, astronomy

Gamma Rays:

Wavelength: < 0.01 nm

Frequency: > 30 EHz

Uses: Cancer treatment, sterilization, nuclear physics”

Make of that as you will, please share your thoughts if you have any before continuing to read.

But for me I think about the full spectrum of colours we can see are obviously very limited. If we were able to perceive all of this spectrum, I image it is just purely more nuance and distinguishing between these levels. So imagine we could distinguish and our perception wasn’t so limited. Radio waves- Gamma rays represents a spectrum from red to violet (red, orange, yellow, green (maybe cyan), blue, (maybe indigo), violet… I notice that Jane has been since the start of the show associated with radio, television. I think this is giving us the key to understand the other characters like Will and Mike… Will has already been hinted at being in control of some characters in a way and I think this is also how the Mind Flayer achieved control of Will by stealing his power and using it against him. Now Mike is interesting because his colour could be UV a symbol for something beyond human perception in a way beyond sight (like emotion - “you’re the heart” OR or X-Rays a possible symbol for the ability to see through something…

what do you think?

[maxs-guesswork]

Unfinished Timeline for an Untitled Setting

Critique and advice is more than welcome, though please be nice about it. Goes up to about 2081 rn, though I plan to get at least another 50 years further in before I get to the time I want the bulk of the setting to be set in.

Timeline:

2022: First controlled break-even fusion reaction, followed by first controlled net-gain fusion reaction.

2025-2026: Increasing unrest in USA leads to mass riots outside the white-house. Sweeping reforms after growing revolts threaten to become a major armed rebellion. NASA miraculously left untouched, general increase in standard of living. Economic crisis narrowly averted.

2027: First nuclear thermal rocket (NTR) tested in orbit by NASA and DARPA. GPT-style language modeling declared “dead end” for self-aware AI.

2030: First Lunar base established under NASA Artemis program. Suez Canal temporarily blocked by a poorly driven cargo ship again. Evergreen Shipping goes bankrupt.

2034: Lunar Gateway established under joint NASA, ESA, JAXA, DLR, ASI, and CNSA. Lunar helium-3 mining declared officially nonviable. Radial detonation engines become standard for lower ascent stages, SpaceX Starship, NASA SLS, and Roscosmos Soyuz phased out. Drop in launch prices.

2034-2036: Additional modules added to the Lunar Gateway from SpaceX, KARI, ISRO, and Roscosmos. Lunar Gateway Collaborative Group (LGCG) established consisting of all current contributors to the station.

2036: First commercial fusion energy plant reaches full operation in France under ITER. Mass production of Tritium begins. First fully private space station under SpaceX. Asteroid mining corporations begin formation. Establishment of Nigerian Organization for the Development of Space (NODS). Ecuador experiences communist revolution.

2036-2037: First manned martian mission under LGCG, first human footsteps on another planetary body.

2037: Elon Musk assassinated. New SpaceX leadership declares plans for space elevator. North Korea collapses, Korean peninsula unified under South Korean leadership, becoming simply Korea. Indian nuclear stockpile secretly surpasses 50000 Gt. First baby born on the moon.

2040: Artemis base becomes semi-self sufficient, producing it’s own food and air from hydroponics, and water from mined lunar ice. Lunar LH2 and hydrolox production begins. Lunar population passes 100.

2040-2042: First commercial fusion power plants established in the US, UK, Australia, Korea, and Japan.

2042: A joint US Government and SpaceX black operation destabilizes Ecuador, leading to a corporate takeover of the territory.

2044: Korea, Japan, the Philippines, Vietnam, Malaysia, Indonesia, Papua New Guinea, Australia, and New Zealand form West Pacific Trade Organization (WPTO). Construction of the base of SpaceX’s planned space elevator begins off the coast of Ecuador.

2047: LCC completed at CERN. Mission for permanent martian base declared. Major economic crisis in China, intervention from several megacorps results in a decrease in Chinese government power and increase in corporate control in the region. SpaceX space elevator counterweight construction begins in geostationary orbit.

2048: Major revolution in quantum mechanics brought on by new data from the LCC. Lunar population passes 250.

2050: China splits into 4 corporate states, Amazon Corporate Territory (ACT) with its capitol in Chongqing, Samsung Independent State (SIS) with its capitol in Shanghai, Territory for Electronic Developments (TED) made up of Apple and Microsoft with its capitol in Yinchuan, and the Chinese Corporate Union (CCU) made up of several formerly state-owned corporations with their capitol in Wuhan and possession of the Three Gorges Dam. Beijing becomes an independent city-state controlled by the former Chinese government, retaining control over the CNSA. Massive revolution in battery energy density. Permanent martian base established by LGCG.

2051: Breakthrough in photon manipulation, beamed energy and solar collection becomes increasingly viable. Many asteroid mining corps branch into solar power, notably Binghamton Vacuum Mining Solutions (BVMS). Lunar population passes 500.

2052: Martian population surpasses 100.

2053: Martian base reaches semi-self sustainability.

2055: All 4 Chinese corporate states and the Beijing city state form the Chinese Federation for Space Exploration (CFSE), supplanting the old CNSA. Lunar Gateway module renamed and LGCG roster amended accordingly. SpaceX space elevator cable completed, first test cart sent to GEO. WPTO begins construction of a space elevator in the Banda Sea.

2056: SpaceX space elevator declared complete, commercial operation begins.

2057: BVMS surpasses $1T in net worth, becomes primary supplier of energy for the Artemis Lunar Base. Lunar Population surpasses 1k, massive migratory population surge begins following influx of energy from BVMS. Martian population surpasses 250. First fusion reactor in Ecuador.

2058: WPTO space elevator counterweight begins construction in GEO.

2060: First fusion reactors in Nigeria and India. First large-scale solar collector on Earth constructed in New York operated by BVMS. Large population surge in Binghamton NY. Lunar population surpasses 5k. Martian space station established. Regulations for GEO development established.

2061: First lunar-built spacecraft flown. Secondary lunar settlement founded by CFSE. Massive influx of funds for the WPTO space elevator from the CFSE, GEO counterweight construction begun. Lunar Gateway population surpasses 100. First fusion reactor in the Democratic Republic of the Congo (DRC), Congo space agency (DRCSA) founded.

2064: WPTO space elevator cable completed, declared complete and opened to commercial operation.

2065: BVME establishes unmanned Mercurian base. CFSE settlement population surpasses 100. Martian population surpasses 500. Lunar Gateway population surpasses 200.

2066: Mass expansion of Artemis Base life support systems using BVMS produced automated construction equipment. Aerostat scientific outpost established by LGCG.

2067: Microbial life discovered on Venus. Venus outpost (and LGCG) acquires substantial funding boost. Artemis base population surpasses 2.5k and begins to plateau.

2069: Unmanned mission to Europa announced by LGCG, plans to use BVMS automated platforms to drill into subsurface ocean established. Martian base purchases automated construction equipment from BVMS, massive population boon ensues. CFSE settlement population surpasses 750. Lunar gateway population surpasses 500. Martian base population surpasses 500. BVME becomes the largest corporate entity in the system.

2070: BVMS performs feasibility study on gas giant aerostat mining platforms.

2071: Study of Venusian lifeforms disproves Earth-Venus panspermia.

2073: BVMS tests laser-sail propulsion on small unmanned craft.

2075: LGCG Europa mission discovers multicellular aquatic life in Europa’s subsurface ocean. Plans for a dedicated research base drafted.

2076: Multi-corporate base established on Ceres to facilitate further asteroid belt mining. BVMS intentionally excluded from this project.

(System effectively split into quarters: Past Venus under BVMS, Between Venus and Mars under LGCG, belt under Multi-corporate mining control, outer system unclaimed.)

2077: GEO-Lunar cycler niche mostly filled by Intraplanetary Transport Services corp (ITS).

2080: Permanent scientific base established at the Europa Breach Point (EBP) with mostly automated systems and a small (5 human) management and maintenance crew.

2081: Panspermia further disproved by study of Europan life. Massive object detected in Jupiter’s lower atmosphere. BVMS begins mission to establish a mining aerostat on Saturn, utilizing laser sail propulsion to transport equipment.

(Saturn Aerostat site intended for use in the further colonization of the outer solar system and the Uranus planetary system itself. Atomic Rockets page)

[ssruis]

cpdb rui + nene cards and emu5 tsukasa + emu cards for the alien alien/niccori survey team/opera space opera “alien race w the goal to spread love and cheer because positive emotions are their society’s fuel (like monsters inc. laughter as a fuel) invasion for Benevolent Reasons” ruikasa/emunene au where emu+tsukasa crash land on earth and are found by rui+nene who have to hide them from authorities/fix their ship… maybe len as the alien ship AI rin as the AI rui made/found/whatever… do u see my vision… the dish I’m cooking up…

Assorted vague ideas that may or may not end up making the cut:

> CPDB cyberpunk dystopia being a result of emu’s grandpa giving earth (his favorite little planet) some technology because he was like surely the humans will use this to usher in a new era of peace and love 😊 (he was wrong)

> emu’s grandpa gave them souped up hyper competent chat gpt/ai to solve their more boring meaningless issues so they could focus on creating art and then capitalism ruined it 😔

> no I don’t have an agenda against ai generation what would make you think that (graphic design major being told to use generative AI because “that’s the future” when it Looks Like Shit)

> emu crash lands them on earth when they’re meant to be going elsewhere because she wanted to take “a quick teeny weeny detour” to look at her grandpa’s favorite planet (she knew something was wrong w it. Energy outputs were bad? Maybe the alien society wrote off earth as a lost cause) and tsukasa’s like no absolutely not (they wrestle with the controls -> crash land -> emu’s like oopsie daisy 😊 while tsukasa has a hysterical fit)

> if I want to be really evil emu’s grandpa could have died on earth and she wanted to see if he was alive? But I don’t want to hurt emu like that… I’d rather just go the path of her Really Really wanting to save earth. Au version of Wonderstage.

> or maybe tsukasa and emu are just sent to see why it’s failing. But I think the crash landing thing is really funny.

> ohh maybe instead of emotions it’s music based (harnessing the true emotions behind music) and the reason earth has begun to fail is the increasing drive to create empty pop songs that go viral. That would be a better explanation than the general positive emotion harvesting if you’re assuming everyone in the CPDB dystopia is “happy” (complacent with the status quo of capitalism driven entertainment)

> “it’s a tragedy to view it from this small fry of a planet, but it’s a comedy if you view it from space” “I fell in love with you from far far away” “when this song reached over there and resonated, it was you who gave it meaning! Sure enough this universe is wonderhoy!”

> “I don’t read the lyrics anyway, so give me something catchy” “is it about the numbers than the artistry/awful price to pay for vitality/then what do we do? Was that how it’s supposed to go?/the mind going in circles and slowly dying away” “don’t disguise yourself, fantasies and ideals are fine”

> “The trembling streetlights, the pelting rain/My wavering feelings, an emotional telepath/The lost two make contact/And my heart learned of love “A fluorescent light repeatedly switching off and on/The paranormal confusion quietly ate into you/The parallel signals intersected due to singular point/Created-reality-blind acceptance-itus, sensitivity instinct”

> “we will set out to study the world’s smiles” “Between us, we have something that will turn the impossible into a possibility”

> ^ I can probably do something with those lyrics. General vibes to work with. Etc.

> (rui kamishiro voice) sometimes the cure for complete apathy/hopelessness and disillusionment with the current state of the world can be two goofy aliens (nene kusanagi voice) what the fuck are you talking about

> obligatory rui making joke threats about dissecting emukasa (as he did w wxs sekai plushies/aliens in 2024 April fools) & tsukasa going “y. You’re not serious right” “hehe” “RIGHT??”

> tsukasa initially is very “we need to fix this ship and Leave we have a job to be doing” (realizes the state of the earth) “nnnrgg FINE we have to fix this”

> (sees one bug) “nvm emu can we explode this planet into a million pieces and leave- emu. Emu why are you letting it crawl on you dONT GET NEAR ME WITH THAT” (emu chasing him around w a centipede in her hands)

> emu is immediately infatuated with nene “your voice is the prettiest thing I’ve heard in the whole universe” -> “you make me happy just by existing and I want to make you smile every day”… nene goes from “I want nothing to do with this” to “ok I’m involved now. Not just because emu asked. Rui stop laughing at me. (Would be crying and throwing up on the inside if emu was even a little upset)”

> as with all my aus the ruikasa dynamic is “oh he’s pretty and very talented” vs “what an interesting guy (I want to put him in a maze and watch him solve it for rewards)” -> “this guy is fucking insane and is going to kill me one day (said with begrudging fondness and irritation at his own shit taste)” vs “what an interesting guy (I want to put him in a maze and watch him solve it for rewards) (also I am madly in love with him)”

> they destroy the ai thing or something yay happy ending. Idk. I’m not a polisci major you can’t expect me to have a good neat solution for a huge societal issue.

> ending is hard because I don’t want to separate them but emu + tsukasa have way too much left on the home planet to ever consider leaving it permanently… hmm ending where emu + tsukasa are like we gotta go back for a bit but we’ll return to check on the progress -> more permanent way for visits to happen is set up with the eventual intention for rui and emu to join tsukasa and emu on their adventures?

> saki as a researcher on the alien planet w the rest of L/N… tsukasa prioritizing getting communication set back up first and foremost and rui’s like “is he gonna contact his home planet and ask to be rescued even after he promised to help” & then tsukasa calls home just to tell saki (+ she updates toya) he’s ok (she was understandably distraught thinking her brother fucking crash landed somewhere and died) tsukasa tenma voice I’ve upset my sister. I’m going to go run directly into traffic.

> rui and nene live together (making money thru mechanics and programming). Nene has a shit load of arcade games that she and rui fixed up. Neither of them like how society is but they’re both kind of like “wtf can we even do” although rui does his guerrilla shows on the side (how he finds emu+tsukasa maybe?). Rui dragging himself home at 4 am wounded from nearly getting caught by security and nene’s like are u fucking kidding me.

> actually I think nene playing video games and making money by being a top player is a funnier idea than her helping rui with mechanics. She has some tech knowledge by virtue of living with him but she’s also rank #1 in several competitive games. Nene is the real bread winner in this household.

> I think nene should play video games with emu… emunene arcade date while ruikasa are in the workshop downstairs fixing up the ship (tsukasa is kinda useless he’s just micromanaging/doesn’t want to third wheel emunene because nene kept giving him death glares last time)

Nene: (long winded gaming explanation) so that’s why you want to keep the higher ground during the game

Emu: woaw! you’re so smart nenechan!!

(Loud explosion heard from downstairs) (tsukasa screaming) (rui laughing)

Nene: um. Is tsukasa ok.

Emu: yep he just does that sometimes 💕!! Ooo what’s this game called??

Nene: (whatever it’s not my problem and I’m having fun with emu) this one is called DDR do you want to play 🥰

[planet-gay-comic]

Cheeky Boys in the Wintergarden Waves of Freedom A Journey through the Evolution of Sexual Morality from Art Nouveau to Late Modernism

The development of sexual morality in the 20th century is a fascinating journey from strict societal norms to an era of personal freedom and self-expression. From Art Nouveau to the vibrant decades of the 60s, 70s, 80s, and 90s, we witnessed an impressive evolution in the perception and practice of sexuality and individual expression.

Art Nouveau - A New Aesthetic Emerges: Art Nouveau not only brought about a change in art and design but also ushered in a relaxation of traditional sexual morals. During this era, there was a growing interest in human sexuality and a subtle challenge to societal taboos, along with a new body consciousness, in an effort to move away from the prudish and rigid Victorian era. A notable feature of Art Nouveau was its association with the Roman Baths, which became a meeting place for people of different sexual orientations. In the Roman Baths, individuals of all sexual preferences came together without inhibition, creating an atmosphere of tolerance and openness. This ambiance contributed to Art Nouveau often being seen as a precursor to the sexual revolution of the 60s.

The 60s and 70s - The Age of the Sexual Revolution: The next decades brought an explosive change in sexual morality in the Western world. The 60s brought a new wave of sexual freedom with the Hippie movement and the introduction of the birth control pill, but that was just the beginning.

In the 70s, growing awareness and increasing acceptance led to a more open attitude toward diverse sexual identities and orientations. This time was marked by the emerging LGBT movement, advocating for the rights and visibility of lesbians, gays, bisexuals, and transgender individuals.

In the 80s and 90s, clubs and discos reached their peak as places of freedom and expression. Iconic songs like "YMCA" by the Village People and "We Are Family" were not only hits but also anthems of the LGBT community, which found a safe and accepting environment in these establishments. Concurrently, the first gay saunas emerged, serving as safe havens and meeting places. The sexual revolution of this era was characterized by an openness and diversity that continues to shape the LGBT movement to this day and represents an important part of the history of sexual liberation.

The evolution of sexual morality in the 20th century tells a story of freedom and progress. From the subtle beginnings of Art Nouveau to the exuberant freedom of the late modern era, these changes reflect the ongoing struggle for personal freedom and the acceptance of different lifestyles.

Text supported by Chat-GPT-4.0 These image is generated with StableDiffusion v1.5. The poses and the backround were obtained with ControlNet v1.1.410 via open pose. Faces and background overworked with composing and inpainting.

[mariacallous]

In late July, OpenAI began rolling out an eerily humanlike voice interface for ChatGPT. In a safety analysis released today, the company acknowledges that this anthropomorphic voice may lure some users into becoming emotionally attached to their chatbot.

The warnings are included in a “system card” for GPT-4o, a technical document that lays out what the company believes are the risks associated with the model, plus details surrounding safety testing and the mitigation efforts the company’s taking to reduce potential risk.

OpenAI has faced scrutiny in recent months after a number of employees working on AI’s long-term risks quit the company. Some subsequently accused OpenAI of taking unnecessary chances and muzzling dissenters in its race to commercialize AI. Revealing more details of OpenAI’s safety regime may help mitigate the criticism and reassure the public that the company takes the issue seriously.

The risks explored in the new system card are wide-ranging, and include the potential for GPT-4o to amplify societal biases, spread disinformation, and aid in the development of chemical or biological weapons. It also discloses details of testing designed to ensure that AI models won’t try to break free of their controls, deceive people, or scheme catastrophic plans.

Some outside experts commend OpenAI for its transparency but say it could go further.

Lucie-Aimée Kaffee, an applied policy researcher at Hugging Face, a company that hosts AI tools, notes that OpenAI's system card for GPT-4o does not include extensive details on the model’s training data or who owns that data. "The question of consent in creating such a large dataset spanning multiple modalities, including text, image, and speech, needs to be addressed," Kaffee says.

Others note that risks could change as tools are used in the wild. “Their internal review should only be the first piece of ensuring AI safety,” says Neil Thompson, a professor at MIT who studies AI risk assessments. “Many risks only manifest when AI is used in the real world. It is important that these other risks are cataloged and evaluated as new models emerge.”

The new system card highlights how rapidly AI risks are evolving with the development of powerful new features such as OpenAI’s voice interface. In May, when the company unveiled its voice mode, which can respond swiftly and handle interruptions in a natural back and forth, many users noticed it appeared overly flirtatious in demos. The company later faced criticism from the actress Scarlett Johansson, who accused it of copying her style of speech.

A section of the system card titled “Anthropomorphization and Emotional Reliance” explores problems that arise when users perceive AI in human terms, something apparently exacerbated by the humanlike voice mode. During the red teaming, or stress testing, of GPT-4o, for instance, OpenAI researchers noticed instances of speech from users that conveyed a sense of emotional connection with the model. For example, people used language such as “This is our last day together.”

Anthropomorphism might cause users to place more trust in the output of a model when it “hallucinates” incorrect information, OpenAI says. Over time, it might even affect users’ relationships with other people. “Users might form social relationships with the AI, reducing their need for human interaction—potentially benefiting lonely individuals but possibly affecting healthy relationships,” the document says.

Joaquin Quiñonero Candela, head of preparedness at OpenAI, says that voice mode could evolve into a uniquely powerful interface. He also notes that the kind of emotional effects seen with GPT-4o can be positive—say, by helping those who are lonely or who need to practice social interactions. He adds that the company will study anthropomorphism and the emotional connections closely, including by monitoring how beta testers interact with ChatGPT. “We don’t have results to share at the moment, but it’s on our list of concerns,” he says.

Other problems arising from voice mode include potential new ways of “jailbreaking” OpenAI’s model—by inputting audio that causes the model to break loose of its restrictions, for instance. The jailbroken voice mode could be coaxed into impersonating a particular person or attempting to read a users’ emotions. The voice mode can also malfunction in response to random noise, OpenAI found, and in one instance, testers noticed it adopting a voice similar to that of the user. OpenAI also says it is studying whether the voice interface might be more effective at persuading people to adopt a particular viewpoint.

OpenAI is not alone in recognizing the risk of AI assistants mimicking human interaction. In April, Google DeepMind released a lengthy paper discussing the potential ethical challenges raised by more capable AI assistants. Iason Gabriel, a staff research scientist at the company and a coauthor of the paper, tells WIRED that chatbots’ ability to use language “creates this impression of genuine intimacy,” adding that he himself had found an experimental voice interface for Google DeepMind’s AI to be especially sticky. “There are all these questions about emotional entanglement,” Gabriel says.

Such emotional ties may be more common than many realize. Some users of chatbots like Character AI and Replika report antisocial tensions resulting from their chat habits. A recent TikTok with almost a million views shows one user apparently so addicted to Character AI that they use the app while watching a movie in a theater. Some commenters mentioned that they would have to be alone to use the chatbot because of the intimacy of their interactions. “I’ll never be on [Character AI] unless I’m in my room,” wrote one.

[abalidoth]

it is fucking wild to me that bad AI (LLM specifically) takes are coming from such vastly separate parts of the conversation

You've got the tech bros on one side going "I talked to chatgpt for an hour and I'm having an existential crisis about the nature of humanity 😨" like my dude. Even on the off chance that this does qualify for sentience (it doesn't) this is something that people much smarter than you have been thinking about for decades. You are not unique in your "but what if AI think DIFFERENT??" tweets.

(...and then these chucklefucks are trying to solve the alignment problem with mind control -- directly injecting contrary thoughts into the internal state of the network. Which if this thing IS as smart as they fear it is, is horrifically unethical, and if it isn't, it's pointless. Sigh.)

But then on the other side you have non-tech anti-AI folks with the "ugh why does anyone think it's intelligent 🙄 it's just mad libs/cut-and-paste prediction" as though GPT is just a really big Markov model or something.

Like, ok, no. The end task is language modeling, yes, but we're asking these systems to model language to such a high precision that the only way to do it is to actually develop an internal conceptual model of the world. This isn't just "this word usually comes after that word", it's "this string of words maps to _this_ concept in thought space, that string maps to this _other_ concept, here's a transform that bridges the two."

Is it conscious? Fuck man I dunno, we kinda have to figure out what the hell consciousness is other than "the thing humans have, you know, the thing where you're conscious!" before we can properly answer that question. But I'd say there's a tremendous amount to support that LLMs have _thoughts_. Because we can see them! They are mostly black boxes, yes, but you can still poke around inside the internal vector spaces and figure out what areas light up with different concepts, same as you can on a human brain with a PET scan.

Yes, the task they're set to is currently language modeling, but the means by which they do it necessarily requires context-dependent abstract concept manipulation on a pretty impressive scale. And that's not bringing in things like AutoGPT where a primary LLM spins up prompts for daughter LLMs in order to solve some specific task. As someone in the field, it's honestly pretty impressive, and we're going to start seeing major shockwaves soon.

This is the first time that the eternal joke of AGI being "twenty years away for seven decades and counting" is starting to feel like it might be outdated. It's gonna change a lot -- don't let the fact that a lot of the voices talking about this are clueless techbro types lull you into thinking otherwise.

[learnwithmearticles]

Technology's Use of Water

While water is renewable, it is finite. Its renewability depends on us using and managing our water resources responsibly.

Previous articles on this page have discussed hydropower and how it produces less waste and costs less than other resources. We have also briefly discussed how other energy sources consume water as a coolant or receptacle for waste. Entire university courses are dedicated to human uses of water.

Water Scarcity

Only 3% of water on Earth is freshwater. Of course, we need this to drink, but we need it for many more services beyond that.

Many plumbing fixtures are made of copper, which saltwater severely corrodes, same as lead and, over a longer time, PVC. Toilets on average use 1-5 gallons of water per flush. If we want to preserve freshwater by switching to saltwater plumbing, we would have to rethink and re-pipe entire plumbing systems.

We lose safe water in rain, as well. Supported by a study in Environmental Science and Technology, the Center for Disease Control and Prevention in 2022 stated that rainwater is not safe to drink. Chemicals known as per-/poly-fluoroalkyl substances break down extremely slowly, and have leached from many products like cleaners, fabrics, and shampoo into the water cycle. Removing PFAS from water requires filters of activated carbon or reverse osmosis membranes, which also require frequent maintenance.

A lot of water is also not available to us because it is in ice caps and glaciers, which are estimated to be about 68% of Earth’s freshwater. This water is also being lost, because as glaciers melt at increasing rates, that freshwater becomes saltwater in the ocean.

These limitations mean that water is not necessarily renewable yet, especially because treating water produces its own waste and pollution. We have to be responsible with the small percentage of water we have access to.

Irresponsible Use

There are a ridiculous amount of ways in which we waste water. Leaks, watering lawns, and leaving taps running are some of the big household wastes of water. While individual accountability and changes can still make a big difference, I want to focus on bigger impacts.

One example is in nuclear power production. Nuclear power plants use water to cool down used fuel when it is done being used in the reactor. This results in radioactive and thermal water pollution.

Agriculture is another common cause of water pollution. Excess water from rain or artificial watering runs off of agricultural fields and flows towards streams and bodies of water. This runoff often includes amounts of fertilizers and pesticides ranging from minimal to extremely harmful. This leads to improper levels of oxygen, nitrogen, and hydrogen within the water. Like water contaminated by pharmaceuticals, this is not safe to drink, and something not safe for skin contact.

Technology is also a major factor of water demands. Artificial Intelligence and cryptocurrency are heavy water consumers.

AI is beneficial within waste management, as it is able to quickly analyze information and identify issues, potential problems, and potential areas of improvement. Unfortunately, AI training requires a large amount of water. One study states that training GPT-3 alone can evaporate 700,000 liters of freshwater. In 2027, AI is predicted to consume 4.2 to 6.6 billion cubic meters of water. In comparison, Denmark nationally consumes around one billion cubic meters in a year.

Cryptocurrency is even worse. It goes through a process called mining in which transactions are verified and new ‘coins’ are generated into the system. This process is extremely water-demanding. For example, in 2021, mining of Bitcoin consumed more than 1,600 gigaliters of global water. On average, each cryptocurrency transaction consumes 16,000 liters of water in cooling down the computer equipment and the power plants that provide the electricity.

Saltwater as an alternative in these situations does exist; however, this process has the disadvantages of one-time use, large water intake, sewage discharge, and ocean pollution. Technology has begun to improve on this method with seawater circulation cooling technology, which reduces sewage discharge and water intake, but remains an imperfect solution.

Technology has the potential to drastically improve environmental management and restoration, but still has a long way to go before we offset the huge impacts we have made. Freshwater is taken for granted by many people, and the systems that disproportionately consume the most of it are not held accountable. This cycle must stop if we want to make water a truly renewable resource.

Additional Resources

1. Water Renewability

2. Corrosion on Plumbing

3. Treating PFAS

4. Household Water Waste

5. Nuclear Water Waste

6. AI Helping Water Management

7. AI Water Consumption

8. Crypto Mining Water Consumption

9. Seawater cooling technology

[thursdayisbetterthanfriday]

Atom: The Beginning & AI Cybersecurity

Atom: The Beginning is a manga about two researchers creating advanced robotic AI systems, such as unit A106. Their breakthrough is the Bewusstein (Translation: awareness) system, which aims to give robots a "heart", or a kind of empathy. In volume 2, A106, or Atom, manages to "beat" the highly advanced robot Mars in a fight using a highly abstracted machine language over WiFi to persuade it to stop.

This may be fiction, but it has parallels with current AI development in the use of specific commands to over-run safety guides. This has been demonstrated in GPT models, such as ChatGPT, where users are able to subvert models to get them to output "banned" information by "pretending" to be another AI system, or other means.

There are parallels to Atom, in a sense with users effectively "persuading" the system to empathise. In reality, this is the consequence of training Large Language Models (LLM's) on relatively un-sorted input data. Until recent guardrail placed by OpenAI there were no commands to "stop" the AI from pretending to be an AI from being a human who COULD perform these actions.

As one research paper put it:

"Such attacks can result in erroneous outputs, model-generated hate speech, and the exposure of users’ sensitive information." Branch, et al. 2022

There are, however, more deliberately malicious actions which AI developers can take to introduce backdoors.

In Atom, Volume 4, Atom faces off against Ivan - a Russian military robot. Ivan, however, has been programmed with data collected from the fight between Mars and Atom.

What the human researchers in the manga didn't realise, was the code transmissions were a kind of highly abstracted machine level conversation. Regardless, the "anti-viral" commands were implemented into Ivan and, as a result, Ivan parrots the words Atom used back to it, causing Atom to deliberately hold back.

In AI cybersecurity terms, this is effectively an AI-on-AI prompt injection attack. Attempting to use the words of the AI against itself to perform malicious acts. Not only can this occur, but AI creators can plant "backdoor commands" into AI systems on creation, where a specific set of inputs can activate functionality hidden to regular users.

This is a key security issue for any company training AI systems, and has led many to reconsider outsourcing AI training of potential high-risk AI systems. Researchers, such as Shafi Goldwasser at UC Berkley are at the cutting edge of this research, doing work compared to the key encryption standards and algorithms research of the 1950s and 60s which have led to today's modern world of highly secure online transactions and messaging services.

From returning database entries, to controlling applied hardware, it is key that these dangers are fully understood on a deep mathematical, logical, basis or else we face the dangerous prospect of future AI systems which can be turned against users.

As AI further develops as a field, these kinds of attacks will need to be prevented, or mitigated against, to ensure the safety of systems that people interact with.

References:

Twitter pranksters derail GPT-3 bot with newly discovered “prompt injection” hack - Ars Technica (16/09/2023)

EVALUATING THE SUSCEPTIBILITY OF PRE-TRAINED LANGUAGE MODELS VIA HANDCRAFTED ADVERSARIAL EXAMPLES - Hezekiah Branch et. al, 2022 Funded by Preamble

In Neural Networks, Unbreakable Locks Can Hide Invisible Doors - Quanta Magazine (02/03/2023)

Planting Undetectable Backdoors in Machine Learning Models - Shafi Goldwasser et.al, UC Berkeley, 2022

[blubberquark]

I could write so many tumblr posts about ChatGPT.

Zero: Already written. GPT-3/Midjourney is not a good tool for procedural level/content generation.

One: Remember when Siri was the future, or when Siri was the beginning of intelligent machines, or when Siri meant humans would just stop thinking for themselves and outsource things to computers, or at least when Siri, Alexa, Cortana, and "OK, Google" were spelling doom for the touchscreen/mouse and keyboard because in The Future, we will all talk to our computers like Captain Picard? Do these people have egg on their face or are they boldly ignoring their past mistakes?

Two: Remember when we called it "Machine Learning" instead of AI, because remember what happened the last time we hyped up things as AI? Why are people doing this again?

Three: Back to Siri. People were prognosticating that Siri would only get smarter. In many ways, it did, but that didn't result in a "general intelligence". And yet, Siri (and "OK Google") knows so many things for sure. Unlike GPT-3, which essentially suffers from fluent aphasia or Korsakoff Syndrome, Siri had a knowledge base and could reason. It wasn't intelligent, I grant you that. But do you understand why Siri, or IBM's Watson, or even Wolfram Alpha did not scale up to become ChatGPT? I mean I do, it's software engineering and marketing and economics of scale. But do those people who make grand predictions about GPT-4 understand this?

Four: Here in Germany, I hear politicians call for a more "competitive" AI policy, which mostly means less data protection. We are already in the absurd situation where a doctor can't publish the success rates of different surgical techniques in retrospect, because that would be a study and subjects have to consent in advance and a study on human subjects needs a good reason and also a control group - while at the same time the government wants to give health data to medical app start-ups in bulk. You think this isn't really about ChatGPT, but it is about machine learning. It looks like the government doesn't want doctors to analyse data, but start-ups, and it doesn't want studies, but products.

Five: AI is a marketing gimmick anyway. Many products just use AI to use AI. Blog posts about using AI to do a task exist to create FOMO in people who don't use AI. Products use "AI" in order to court controversy.

Six: Prompt injection and prompt leaking should be easily solved in principle, and I am sure by this time next year they will be "solved", and have been in some proof of concept projects, but in practice economic incentives apply that make this difficult or we would have solved it already.

Seven: Prompt engineering is difficult. It requires some insight into the behaviour of a language model, or at least its inner workings. Will prompt engineering stay relevant? On a related note, Google-Fu still as relevant as it used to be in 2004?

Eight: Did Siri get worse?

[cbirt]

OpenAI’s ChatGPT: A Promising AI Tool for Early Detection of Alzheimer’s Disease

Researchers at Drexel University’s School of Biomedical Engineering, Science, and Health Systems have found that ChatGPT, an AI-powered chatbot program, may be able to detect early signs of Alzheimer’s disease. In a study, the program analyzed conversations between individuals with Alzheimer’s disease and healthy controls and was able to accurately identify those with the disease with an accuracy rate of 80%. ChatGPT uses AI algorithms to generate responses similar to how a human would respond to user input and be able to detect differences in language patterns, such as difficulty finding the right words and repetitive language, in individuals with Alzheimer’s disease. While more research is needed, the study suggests that OpenAI’s GPT-3 could potentially be used as a non-invasive, cost-effective method for detecting Alzheimer’s disease in its early stages. 

Continue Reading