#My material is not extracted from any one specific source

[deathinfeathers]

.

[soylent-crocodile]

Gourmand (Alchemist Archetype)

(art by Htg17 on deviantart)

(If you know me this class is... not entirely surprising, but it sure is different! The idea of a class that temporarily steals abilities from monsters is a tricky one to balance, and I haven't got very far in my playtesting for this, so... handle with care!)

There are many reasons for adventuring; wealth, glory, or even just escaping your awful hometown. Among the oddest, though, is the motivation of the gourmand; the desire to taste. There are thousands of monsters out there, from carnivorous plants to magical animals to even the mighty dragon, and the gourmand’s desire is to taste them all- and gain a temporary modicum of their power!

Note that the gourmand has no restriction on the consumption of humanoids and other sentient creatures- but also note that such acts are almost universally frowned upon or even illegal in all but the most brutal and violent societies.

Delicacy

Rather than brew mutagens, a gourmand crafts perfectly made delicacies. These delicacies are specifically made for her palate, as well as for her unusual constitution gained by years of exotic meals; a creature other than the gourmand who eats it simply finds it a delicious meal and gains no additional effects. Creating a delicacy requires an hour of time, as well as at least a half pound of organic materials. These materials can be as simple as plant matter harvested from the area or farm raised meat, but gaining access to rarer meats provides additional effects. This matter, however, must have been part of a living creature within the last week; beyond that, it has lost its freshness, despite any magical or mundane preservation effects it may have been treated with. Once created, a delicacy must be consumed within a day or its effects are lost. Note that some meats are poisonous, and that it sometimes may be impossible to remove the poison from the food; a gourmand must rely on the protection from poison granted by her alchemist abilities to overcome this. A gourmand can only be under the effect of one delicacy at a time. A delicacy is mutually exclusive with a mutagen or cognatogen; the taste of a delicacy is ruined by the presence of such elixirs in the gourmand’s body. The effects of delicacy lasts 10 minutes per level in gourmand, and grats a +1 alchemical bonus to attack rolls, damage rolls, and fortitude saves, and 1 temporary hit point per HD. However, she gains a -1 penalty to reflex saves, as she is feeling a bit stuffed. Additionally, the gourmand may gain one of the following effects, based on any organic creature that is used as part of the recipe, henceforth referred to as ingredients;

Might- Access to one combat feat an ingredient had. The gourmand must match the prerequisites of that feat to use it, but treats her level in gourmand as levels in a full BAB class for this purpose.

Magic- Access to one spell or spell-like ability an ingredient had. The spell or spell-like ability must be of equal or lesser level to the highest alchemist extract the gourmand can cast. If the spell level is of the highest alchemist extract the gourmand can cast, she can cast it 1/day. Otherwise, she can cast it 3/day. This ability uses her intelligence modifier, rather than her charisma, to determine the DC.

Protection- Resistance 10 to one or fewer elements an ingredient is immune or resistant to.

Speed- A +10 bonus to her movement speed, provided an ingredient has a faster movement speed than her or a movement type she does not have.

This ability replaces Mutagen

Sneak Attack

A gourmand prefers to disassemble her victims, rather than blowing them to (overcooked) smithereens. She gains the sneak attack ability as a rogue of the same level. If she has sneak attack from another source, this ability stacks with it.

This ability replaces Bombs.

Scent

A gourmand’s nose is finely tuned to the variety of tastes and smells that surround her. She gains the scent ability, as the universal monster rule.

Bite

At level 2, a gourmand gains a bite attack as appropriate for a creature her size if she does not already have one. If she is wielding a weapon, this is a secondary natural attack. At level 6, the bite gains the grab ability, but the alchemist takes a -4 penalty to grapple checks made with this ability. If she already has a bite attack, she gains grab.

This ability replaces poison use and swift poisoning.

Discoveries

A gourmand gains access to the following discoveries in addition to those gained by normal alchemists.

Greater Delicacy Prerequisites: Alchemist 12 A gourmand with this discovery gets a +2 alchemical bonus to attack rolls, damage rolls, and fortitude saves and 2 temporary hit points per level, rather than the 1 naturally granted. Additionally, she may pick two effects granted by her delicacy, rather than one, and takes a -2 penalty to reflex saves.

Grand Delicacy Prerequisites: Alchemist 16, Greater Delicacy A gourmand with this discovery gets a +3 alchemical bonus to attack rolls, damage rolls, and fortitude saves and 3 temporary hit points per level, rather than the 1 naturally generated. Additionally, she may pick three effects granted by her delicacy, rather than one, and takes a -3 penalty to reflex saves.

Taste for Combat Prerequisites: Alchemist 8 When a gourmand with this discovery selects the Might effect for her delicacy, she gains access to up to two combat feats an ingredient had, rather than the one normally granted. These need not be from the same ingredients.

Taste for Magic Prerequisites: Alchemist 8 When a gourmand with this discovery selects the Magic effect with her delicacy, she may gain access to up to two spells or spell-like abilities possessed by ingredients, rather than the one normally granted. These need not be from the same ingredients.

Used to the Spice Prerequisites: Alchemist 10 When a gourmand with this discovery selects the Protection effect with her delicacy, she instead gains immunity to that element, even if the ingredient only had resistance.

Surf and Turf Prerequisites: Alchemist 6 When a gourmand with this discovery selects the Movement effect with her delicacy, she additionally gains her choice of a fly speed (double her land speed, clumsy), swim speed (equal to her land speed), or climb speed (equal to her land speed), provided that an ingredient has that movement speed. If she gains a swim speed, she also gains the ability to breathe underwater.

Magically Delicious When a gourmand with this discovery selects the Protection effect with her delicacy, she also gains SR equal to 7+her level in gourmand. The SR need not be from the same ingredient as the elemental resistance, if any.

Gastrolith Prerequisites: Alchemist 8 When a gourmand with this discovery selects the Protection effect with her delicacy, she gains DR2 that is overcome the same way as the DR of an ingredient. This need not be the same ingredient as the elemental resistance, if any.

Lockjaw Prerequisites: Alchemist 6 A gourmand with this discovery ignores the -4 penalty to grapple checks made with her bite’s grab ability and increases the damage dice for her bite as though she were a size larger.

Voracious Prerequisites: Lockjaw, Alchemist 10 A gourmand with this ability gains the swallow whole universal monster ability. She may use this ability on any creature up to her size category. This ability deals 1d6 damage each round for every 3 levels of Gourmand, the damage being half acid and half bludgeoning. A creature eaten this way has not undergone the careful preparations required to turn it into a delicacy and as such the gourmand gains no other benefits from it.

Overstuffed

At level 16, a gourmand’s delicacy lasts 1 hour/level.

This ability replaces Persistent Mutagen.

Grand Discovery

A gourmand gains access to the following grand discoveries in addition to the ones normally available.

True Delicacy Prerequisites: Grand Discovery, Grand Delicacy A gourmand with this discovery gets a +4 alchemical bonus to attack rolls, damage rolls, and fortitude saves and 4 temporary hit points per level, rather than the 1 naturally granted. Additionally, she may pick any number of effects granted by her delicacy, rather than one, and takes a -4 penalty to reflex saves.

Incredibly Voracious Prerequisites: Grand Discovery, Voracious A gourmand with this discovery gains a permanent +4 bonus to grapple checks, the Fast Swallow ability as the universal monster rule, and their swallow whole ability deals 1d6 damage for every two levels in gourmand

[alatismeni-theitsa]

Hi!

I have heard from a lot of people that the churches built over pagan temples had a pretty standard correspondence between the saint the church was devoted to and the God that used to be worshipped there. I have tried looking it up on the internet but, as you know, Greek sources are hard to come by and I haven't found anything in English either. Any chance you've read/heard anything on the subject?

Unfortunately, I cannot be too specific here, but I'll do my best. "Build over" it's not the correct description in most cases, I think. The usual situation is: because the site was already deemed sacred and the old faith was dying or had died already, materials were taken from the old temple to build a new one pretty close. I remember such a thing happening with a temple of Aphrodite in Cyprus.

But I heard of many cases over the years. It's such a usual phenomenon that is pretty well known by many Greeks. The knowledge is in the public consciousness to the point that we don't need to prove it to ourselves. (Of course there is plenty of evidence, but the average Greek can hear this info pretty easily)

In some cases, the old temples were straight up reused. The Parthenon itself became a temple to the Panagia (Virgin Mary) and it was a Christian building for as long as it was a Polytheistic one. And then the Ottomans turned it into a mosque, and then a storage for explosives (that didn't go well).

Materials were re-used all the time in the Mediterranean region, not just by Greeks but by any nation in those areas and those who conquered them, like the Ottomans. Then, the Greeks and other peoples used the Ottoman materials and buildings when the Ottomans left. We modern people often forget how effort and time-consuming it was for people to build anything back in the day. Having a building or parts of a building was a great advantage that people used, because the average person back then had very little means available. Or, they had to already rebuild everything after they got conquered, or after they conquered others.

Sometimes western Hellenic polytheists get shocked and offended that buildings were reused but, babes, that is what people have been doing for thousands of years and in many, many faiths. When the population of a faith is down to very few people, what will the goat herders and farmers be supposed to do? Get the funding, extract marble from kilometers away, and build something from the ground up, while letting the old building rot, while living in a very seismogenic area that already destroyed their buildings every few generations. Greeks and their neighbors were always practical people, and they had a hard time surviving already. It's common sense that you will reuse buildings and materials.

If you want to begin from somewhere you can start with the Parthenon, and the connection between the Cypriot Mother Goddess Aphrodite and the Panagia (Virgin Mary) and their sacred sites. Additionally, you can take a look on a paper I recently stumbled upon, which is indicative of the situation. "The Reuse Of Marble In The Eastern Mediterranean In Medieval Times" By Godfrey Goodwin.

[vraisetzen]

Wonderful V,

I honestly think you’re one of the most interesting writers on this platform and currently, you’re the only one whose blog I am interested it and regularly check up on.

You are such a phenomenally awesome writer. Because I adore your writing, I ended up reading your one-shot for ‘The Elusive Samurai’, even though I haven’t read the manga or watched the anime.

I really, really enjoyed it. I still remember it in my mind and if asked, I would be able to talk about it still! In my mind, it feels very cinematic. I was fully engrossed in your capturing of their world!

I honestly feel like your attention to detail and your effort to remain historically accurate mixed in with your natural talent for choosing some of the most visually evocative and linguistically stunning imagery makes you an extremely compelling writer. I really enjoy reading all your works, regardless of whether I’ve seen the source material or not.

Because you are such a talented writer, I am not sure what to request! The way you write both Kokushibo and Douma is brilliant. I won’t lie — I really adore the way in which you write Kokushibo and I think you write him so well that you almost do him more justice than the source material. But your Douma is also so phenomenally in character!

I haven’t read your Muzan, but with how well you tackle literally every single character from ‘Demon Slayer’, I’m sure you’d be able to do him justice.

I remember reading a while ago your thoughts on Kimetsu Academy, so I feel like it’d be fun to request a modern set fic with either Kokushibo or Muzan.

But the thing is, your attention to historical accuracy is so phenomenal that I’m really tempted to request a historical fic. I honestly think that out of every fan fic I’ve ever read, you’re the only one who writes like a novelist and not a fan fic writer. That is to say, you put in so much effort into historical accuracy that you almost transform yourself into a historian and I really admire that.

You have the ability to be able to write imagery well, to be able to plot well, to have the imagination to come up with creative plots, and to have the capacity and determination to remain as historically accurate as possible.

I’m honestly really impressed by your writing abilities. Seeing your writing really makes me smile and reading your accompanying little ficlets for ‘Notte Stellata’ became the sole reason why I’m still even checking Tumblr. Your writing honestly makes my day and I get really excited whenever I see that you decided to write a little extract or a snippet or a fic.

I’m sorry for the ramble and I’m sorry I wasn’t more specific in my request! The thing is, I like your writing so much that I’d be happy with almost any request! (Though I do want to say… I’d prefer to request a semi-longform over a drabble!)

Hi Anon! Thank you for your very kind words — I'm so happy that you enjoyed the Yorishige oneshot even though you hadn't seen The Elusive Samurai (although I highly recommend it!)

Likewise, I'm really glad that you found my fics to be historically accurate! It means a lot to me that the setting of my story, the objects I reference, and even the characters' speech are true to the cultural mores of that time period — even if its a series like The Elusive Samurai which has a very playful, modern sense of humour. I do think staying true to the historical setting helps with establishing the atmosphere of the story — and with that comes with the respective imagery that I wish to evoke. Of course, being a history nerd myself, I don't mind doing a little bit of research :D

And well noted also on your request! I have gotten a few on a modern AU, and I'm deciding between the Kimetsu Gakuen AU or something completely different; we'll cross the bridge when it gets there, as they say!

I do apologise for my late response too — I've received so many wonderful messages in the past week and I've only recently had the chance to look at them. I always value everyone's kind reviews (and criticism, if any) and I genuinely feel so loved by the response I'm getting to my fics.

Thank you again!

xoxo, V ♥️

[sleepysadden]

My gripes with AI and its acceptance

People on the internet participating in spaces centred around creative work of any kind (both technical and more free-form) are often wary of infringing on creative rights of others. These people are an important factor in enforcing proper treatment of original authors by pointing out cropped references to the author, blurred watermarks, providing sources and meticulously analysing code they suspect to be stolen.

It's always satisfying to see a person trying to pass someone's work as their one be called out for it and made to stop through pure societal pressure directed towards something deemed unjust. Unfortunately these techniques are applicable only on singular individuals or small groups. Any bigger entity tends to lack a 'face' towards which any attempts can be directed. This is why its often much harder to enforce rights of smaller creatives whose work is being stolen by bigger entities. It requires an official process often going through channels of the entity in question (dubious effectiveness) or through the legal system (expensive, tasking, and long). And that is if the copyright infringement could be shown, which used to be somewhat easy. AI changed that.

Casey Rickey demonstrated on his YouTube channel how one could go about replicating a specific style using his own and Salvador Dali's work as an example. The results were in my opinion not exactly successful but close enough to be concerning. Currently available AI models are cheap for what they can produce and while their results are sub-par and often recognisable as AI (which is also changing rapidly) they are free or close to free which is much less than an artist would charge for a commission. This is likely enough to undermine real artists through: 1 - their works likely being already included in current models, 2 - ability to copy their exact style to a text or image prompt. Both of those currently seem to be legally untouchable as in both cases unlawful (and unethical) use of source material is near impossible to prove outside of catching the culprit in-act, as mentions of similar style could be dismissed as subjective and not exactly proving anything (which is fair, but doesn't help the victim).

Is AI generated content derivative?

We could assume AI generated content to be a more advanced way of taking inspiration in someone's work and trying to replicate things that we deem worth replicating in it, an admiration-through-imitation kind of approach. Though I would argue that intent here is important. Is the intent really to appreciate someone's work or is it to gain a close-enough copy of the work? In my case I consider it more as a form of advanced tracing, which doesn't gel well with being derivative.

Being more on the conceptual side - AI, in it's current, non-consensual form is theft. But what is it stealing? It's not exactly the works themselves, it's more (at the risk of being too abstract) about the work that was put in. AI is not a way to steal the effect of someone's work, but rather to steal days of training and personal input that made the work the way it is. It's offering to apply someone's expertise to an idea of a stranger at no cost to them and with potential loss of income by the original artist. It's a commission, but the artist is absent.

The above was written without aid of generative AI's. I might have more to say about this later, but it's enough for now.

Edit: I forgot about my actual point. My conclusion is that because AI generated content in its current form is inseparably involved in copyright infringement of people whose works cannot be extracted from the model nor can they be properly compensated for damages (as it is almost impossible to say for certain whether their rights are infringed upon or not) - all authors of generative AI's that cannot provide a complete list of sources for its training data along with proof of rights to using them cannot offer their AI as a service, cannot accept payment for using them, should not have created them and may not create any without providing such sources.

[shukuchiisms]

I logged in via phone today and just saw a Reply to my post about kumihos having both fox and human ears, made by @icykalisartblog and I would like to address some points made by the text.

While I don't believe the poster had any intentions to insinuate I was lying or something like that. we all know very well how the internet folk tm works, so let me clarify first and foremost that I never had any intention to spread false information and that I was just using the material I own to talk about my passion for those myths. Now:

[TL DR: Thanks for pointing those out and sorry for my mistakes, but it wasn't intentional in any way, since this one would look gigantic if I just replied on the original I gonna let this here]

The poster claims [and with some solid evidence] that the art I posted from the book Wunwudocheop (or my PT BR translated version at least) is, in fact, edited art from paintings by Shin Yuk-bok, AKA Hyewon, a satire/erotica painter from the Joseon era. And being honest, it is no surprise that they might be in fact, some of my sources had been incorrect in the past and might be so again, but regardless, those pics I posted are extracted from the version of a Joseon dynasty art book that I physically own.

So, just to reiterate, I had no intention to post misinformation of any kind, and I trusted the sources I own. Since I don't possess knowledge of the Korean tongue, I can't verify the correctness of those translations myself. Because like I said before, we all know how the internet works. and I apologize for committing such a mistake, and that at no point I alternated those images in any form.

Now, addressing these tags on the post in specific.

While I don't have the book Yojeon, Dokkaebi, Kumiho & Yulyeong" by Im bang & Yi Ryuk physically with me at moment, I have the Portuguese version in pdf, and it does have the exact the same text word by word I posted on the original, but in Portuguese of the "The salt seller and the fox" tale, as you can see here:

And while I don't doubt that could be discrepancies between the versions, I have no evidence that proves either translation is more correct than the other. Aside from the poster's insinuation that none of the books owned by them possess any tale mentioning that they have fox ears or tails, which gives no indication of which one is the most correctly translated material...(While if I had to guess, the Portuguese translations might have confused different versions of the tale and attributed them to a more famous author by mistake. Since it's not unheard of).

Nonetheless, Mayer, Frederick J still quotes this exact version of the tale in his own works, thus proving it was not a completely fabricated one.

The book Myths and Legends from Korea: An Annotated Compendium of Ancient and Modern Tales. [PDF version] Also has a direct mention of the exactly same tale:

The Mythica encyclopedia also brings the following text:

We can also consult the "The Tails of the Kumiho" text by Tayida Phanich:

So, while you do have a point and some of my sources might be wrong and even invalid in their points, the whole "Kumihos do have both fox and human ears" is still validated by other sources.

Anyway, I will go back and edit the original post when I get the time to remove the invalid sources and add new ones that still affirm the subject as being true.

[whatudottu]

Me, remembering that pill bugs can filter heavy metals from soil:

My brain: You know what this means, don’t you?

Me: What-

My brain: You can make talpaedans part pill bug-

Me, working on amperi headcanons: Wait LET ME FINI-

I’ve gone over a few times that I like to think that talpaedans are techno-organic, less so being silicon or silicone based beings but being an integration of organic carbon materials and the more industrialised physiology of being a walking talking construction equipment. I have also previously made use of the whole ‘machine’ aspect of their natural biology to not only consider them as burrowers, but also as ant-like burrowers with haplodiplody, this form of sex-determination system being used to expand the combinations of construction equipment that an individual talpaedan might have, but also as a basis of which I form the sociopolitical building of talpaedan cities, in which case we have large city state colonies of ant inspired pangolins recreating feudalism Europe of the 10th-13th century.

But up - until this point - there was not quite a lot of circumstances that would allow for actual metallic components to not only present itself biologically but also do so naturally, so with the heavy metal eating pill bug there’s a source of extra inspiration. Oniscidae (apparently the more accurate name for pill bugs/rollie pollies and literally any name under the sun) can remove arsenic (which is actually a metalloid), cadmium, lead, and depending on sources also mercury from the soil; additional metals that can be extracted from soil are copper and zinc. The pill bug in question would crystallise these ions in their gut and create spherical deposits within them, but since we are not talking about the isopod alone surely it isn’t too much of a stretch to the idea that perhaps talpaedans can use these deposits within themselves to feed into the development of their most notable features, their built in equipment?

Assuming that Poiana Lüncas has these metals and that talpaedans would in fact filter for them, in real world human employed Earth uses of these metals can indeed be of fine use. Both arsenic and cadmium can be used as alloys and zinc specifically can galvanise other metals such as iron to prevent rusting, very significant indeed in industry, notably especially with large amounts of zinc can be used in hardware industries. Copper has a very notable use in electronics and wiring of which motors are included, let alone more refined wiring in TVs and radio. And argueably the more infamous of these metals (at least to a standard ‘basic’ level of understanding) we have mercury and lead, the former being used in many different industries for it’s ability to measure the change in temperature and pressure, and the latter for some solders, gasoline/petroleum and wonderfully hazardous cosmetic items that uninformed humanity has come to be harmed by.

ngl, spent a bit of time trying to find an oxygen-carrier of these metals that can feasibly make black blood but i literally do not know if the colour of oxides is in any way relevant to how oxygen-carrier bloodcells would actually be coloured as i’m not a chemist

But in highlighting these metals, I am not simply looking to see if a techno-organic talpaedan can have galvanised armour nor if they come built in with blood pressure monitors. No no no, I am saying more than what talpaedans eat but also what metals can talpaedans have easy access to in the development of technology, and how one might say that actually interacts with the ant-psychology of utterly fucking despising other colony.

Well-

Let’s say that google might be a little concerned about me learning about elements used in the construction of ammunition and radiation protection ehek-!

With the earth tilling that many developing talpaedans need to incorporate necessary metals into their diet, depending on the availabilty of said metals they may be an over abundance for food alone; especially with the pressure of competition of resources in combination with a general disdain for anything extracolonial, a factor that may lead to an escalation of tech development. There is an inherent baseline for technology in talpaedans that would already give their industrialisation a boost, in addition to the materials that may vary on quantity based on the region, colony size, and general population needs. Before long there are city states reinforcing their own barriers and expanding their borders for more and more materials - to consume and to create - to in fact increase and increase the tension between colonies so loud to the point pressure spills like a broken thermometer spills mercury and conflicts arise.

And when everything boils over do they realise that many have to balance the act of raising young and building more weapons, the factor that may in fact be the tipping point to realise that the colonies with the most resources are not only a threat but a threat that can overwhelm on the technological scale and the population number. In spite of how many colonies may indeed hate each other, it is more than worth it to them to compromise and in fact LIVE rather than be caught between the suddenly unveiling superpowers of the world that had previously lay unknown thanks to willful ignorance. Decimation of land is significantly reduced from cataclysmic to catastrophic as entire colonies are rased and in fact dug out into craters, many war-era alliances held to the modern day of Poiana Lüncas. A tithing - remnants of the treaties formed at the peak of desperation - of the youngest generation of adult men stands as both tradition and of appeasement, a colony trade and marriage across many different colonial alliances as a symbol of what had to be done before, sharing resources and rearing while the soldiers fought to live.

And all of this because they got a pill bug diet see this is why ants aren’t allowed to have materials for megaton bombs smh 😞

[snickerdoodlles]

Don’t you feel that if you go on the “loves ao3 more than their own mom” website aka the “can't throw a rock without hitting a writer” website and put in the main tag the opinion that AI writing models are no big deal and don’t do any harm and nobody has any reason to feel hurt over their unethical training in a tone that distinctly resembles the “toughen up snowflake” rhetoric, you might, in fact, just be the one out of line and also an asshole?

*rolls eyes* not what i did nor what i said.

there are many concerns to have with generative AI, such as the ability to extract privatized information from their training datasets, the exploitation of human workers for AI services (tho this one frankly goes for all internet services, not just AI), AI's reinforced biases and lack of learning, and the current lack of regulation against AI developers and AI usage to name a few. in terms of a direct impact on the creative industry, there are several concerns about the uncompensated and unregulated use of copyrighted materials in training data (paper discussing BookCorpus, courtlistener link for writers suing over Books2), the even worse image scraping for diffusion models, screen production companies trying to pressure people into selling their personal image rights for AI use, and publishers getting slammed with various AI generated content while the copyright laws for it are still massively in flux.

i said fanfic does not intersect with AI. actually, i vaguely whined about it in the tags of an untagged post, because i'm allowed to do that on my personal whine-into-the-void space. (which, tumblr is bad about filtering properly in tags and i'm sorry if it popped up anyways, but i also can't control tumblr search not functioning properly.)

there are concerns to be had about AI training datasets (developers refusing to remove or protect private information because it weakens the training data even tho this is a bigger issue for bigger models is my primary concern personally, but the book shadow libraries and mass image scraping are shady ass shit too). but AO3 was never used to train AI. there is a lot of sketchiness involved with AI training data, but AO3 is not one of them.

i get irked when people compare AI generated writing to AI generated art, because the technology behind it is different. to make art, AI has to directly use the source image to create the final output. this is why people can reverse the process on AI art models to extract the source images. written models (LLMs) learn how to string words into sentences and in terms of remembering the specific training data, LLMs actually have a known issue of wandering attention for general written training material like books/articles/etc. (re the writers' lawsuit -- we know AI developers are pulling shady shit with their use of books, AI developers know they're pulling shady shit with their use of books, but unfortunately the specific proof the writers' are using for their case very closely resembles the summaries and written reviews on their books' wikipedia pages. the burden of proof for copyright violation is really hard to prove for books, esp because copyright protects the expression of an idea, not the idea or individual sentences of a work, and LLMs do not retain their written training material in that way.) these are different issues that can't truly be conflated because the methods in which the materials are used and the potential regulation/the impact of potential regulation on them are different.

anyways, back to my annoyances with fanfic x AI -- fanfic is not involved in its development, and if you don't want to read fic made by AI, don't click on fic that involves AI. ultimately, if you read a fic and it turns out AI was involved...nothing happens. if you don't like it, you click a back button, delete a bookmark, and/or mute a user. AI just strings words together. that's it. acting like AI will have a great impact on fandom, or that fandom will be some final bastion against it, is really fucking annoying to me because fandom does not have any stakes in this. there are legitimate issues in regards to developing and regulating AI (link to the US senate hearing again, because there are so many), but "oh no, what if i read a fic written by AI" is a rather tone fucking deaf complaint, don't you think?

[femchef]

So today was the first day back for teachers (semester starts on Monday) and I was going over my notes from one of my classes and picking out things to share with another teacher who’s teaching the same class on one of our satellite campuses -

So anyway, does anyone want to read my lecture notes on thickening agents that I turned into a study guide (I wrote one up during the semester because we didn’t have any previously prepared materials because other instructors just skip or gloss over the chapter but I felt like it was worth the time to focus on the topic and also I’m a Nerd about stuff that makes things gooey).

You know what - I’m just going to post it under a cut below, because it’s fun and also an infodump.

**For context, even though my notes go into more specifics than the required reading, the book for the class is called How Baking Works by Paula Figoni (3rd edition, tbh needs an update but is a good reference), and the link I am telling them to refer to for more information on gelatin that includes conversion charts is here:

Thickening Agents Study Guide

1. Thickening vs. Gelling

1.1. Thickening = moving slowly, viscous, but still some movement while set

1.1.1. Either when sugars and proteins become loosely entangled or when water is absorbed and trapped by swollen starch granules, or when air bubbles in foams or fat droplets in an emulsion slow water movement.

1.2. Gelling = completely set, no movement whatsoever

1.2.1. When water and other molecules are prevented from moving around at all, usually when sugars and proteins bond or tightly entangle and form a larger network that entraps water and other molecules.

1.3. A number of thickening/gelling agents are interchangeable in different quantities.

2. Food-Grade Gelatin (Type A Gelatin) is produced by boiling or soaking pigskins in acid; the connective tissue breaks down into thick strands of collagen and thinner strands of gelatin

3. Powdered Gelatin is made from lower-grade pulverized sheets

4. For more information, take some time to view the attached link in blackboard and the conversion charts.

5. Vegetable Gums = polysaccharides that absorb large quantities of water and swell to produce thick liquids and gels. Veg Gums are a nice source of dietary fiber (think fiber one Powder added to drinks)

5.1. Pectin = present in all fruits

5.1.1. LM (low Methoxyl) Pectin = Also comes from citrus peels or apple. Used in low-calorie jams and jellies, relies on calcium rather than sugar to solidify. Suitable for dairy-based products. Becomes increasingly firm as calcium is added until it reaches saturation point, at which time it begins to reverse in process and soften.

5.1.2. HM (high Methoxyl) Pectin = Comes as Rapid Set or Slow Set; extracted from citrus fruit peels. Rapid-Set for products that require suspension; Slow-Set for recipes that require a smooth texture with no suspension (such as a jelly)

5.1.3. NH (Thermal Reversible) Pectin = Modified LMP; Requires sugar and acidity to gel (and less calcium), and can be melted, set and remelted – requires heat to activate properly. ‘NH’ because of the Ammonia Hydroxide treatment it receives to modify (NH3(aq))

5.1.4. Apple Pectin = Derived from apples. Usually sold as a powder, can be used as a gelling and thickening agent, as well as a stabilizer. Is high in healthy carbs, dietary fiber, sodium, manganese, copper, and zinc – which is why it is a common ingredient in health supplements and pharmaceuticals. Additionally used in laxatives for natural purgative qualities.

5.2. Agar = Is a polysaccharide extracted from either of two varieties of red algae (ogonori and tengusa); has gelling/setting properties that behave remarkably like animal protein. Less agar is required than gelatin, and agar has the benefit of holding shape at room temperature. Cannot be used to stabilize aerated products, and does not whip well.

5.3. Carrageenan = a family of sulphated polysaccharides, name comes from variety of red seaweed found off the Irish Coast termed “Irish Moss”. Typically used in conjunction with meat and dairy products, for which they work particularly well, in large-scale production for stabilization, thickening gelling and texturing.

5.4. Guar and Locust Bean Gum

5.4.1. Guar Gum = Extracted from the endosperm of Guar Beans (legume); does not self-gel like LBG, but is more soluble. Requires high temperatures, high ph and longer times to cause gelling. Low-cost alternative to many other agents and starches, and is 8 times more effective than cornstarch. Used commercially, and stays stable when frozen/thawed.

5.4.2. Locust Bean Gum = Extracted from endosperm of bean on Carob Tree. Dispersible in hot and cold liquid, and converts to gel with addition of minimal amount of sodium borate. Is naturally sweet and is typically used to sweeten foods and as a replacement for chocolate.

5.5. Gum Arabic = Acacia/Senegal/Indian/Sudani Gum = Harvested from Sap of two Acacia Tree Species. Primarily used as a stabilizer (such as in sodas and cosmetics).

5.6. Gum Tragacanth = derived from several species of legumes in the genus Astragalus (Tragacanth, lit. “Goat + Thorn”, which is common name). Largely produced/exported from Iran. Is viscous, odorless, and tasteless water-soluble sap. Traditional binder for pigments in artist’s pastels, and main gum used in fabricated Gumpaste.

5.7. Xanthan Gum = derived from a species of bacteria, Xanthomonas Campestris (same bacteria which causes a variety of plant diseases, such as black rot in brassicas and bacterial wilt in turf grass). Produced via fermentation of glucose and sucrose. Is used to stabilize emulsions (is not an emulsifier in itself). Also helps suspend solid particles in liquids. Commonly used as a thickener in egg white substitutes and to build matrix in gluten-free products where there is no gluten-development.

5.7.1. Shear Thinning/Pseudo-Plasticity: Non-Newtonian behavior of fluids who’s viscosity decreases under ‘shear strain’. Examples Ketchup and Salad Dressing.

5.8. Methylcellulose = “Modified Vegetable Gum” an emulsifier and bulk-forming laxative. Unique property of Setting when Hot and Melting when Cold – commonly used in ice creams for this reason.

6. Starches = Starch molecules are polysaccharides that are arranged in one of 2 ways: either as long, straight chains or as short, but highly branched chains.

6.1. Amylose = long, straight chain starches

6.1.1.  Clouds when cooled

6.1.2.   Firm, heavy-bodied gel when cooled

6.1.3.   Not freezer stable

6.1.4.   Thicker cold than Hot

6.1.5.  Masks flavors

6.2. Amylopectin = short, branched chain starches

6.2.1.   High Clarity

6.2.2.   Thickens, but does not Gel

6.2.3.   Less Likely to weep over time

6.2.4.  Less likely to weep when thawed (more freezer-stable)

6.2.5.  Same thickness hot or cold

6.2.6.  Less likely to mask flavors

6.3. Cereal Starches = extracted from endosperm of cereal grains

6.3.1. Cornstarch

6.3.2. Rice Starch

6.3.3. Wheat Starch

6.3.4. Waxy Maize

6.4. Root Starches = Extracted from roots/tuber plants

6.4.1. Potato Starch

6.4.2. Tapioca Starch

6.5. Modified Food Starches = Starches treated with one or more chemicals to possess more desirable properties or results. (i.e. increased stability with excessive heat/acid, texture, speed of setting)

6.5.1. Corn

6.5.2. Potato

6.5.3. Arrowroot

6.5.4. Tapioca

6.5.5. Waxy Maize (clear and clearer tasting)

6.5.6. Instant Starches = pregelatinized or cold-water swelling (jello cold pudding mix).

6.6. Refer back to previous chapters about gelatinization of starches

6.7. Refer to chart 12.5, pg. 337 for a comparison of properties

Homework: 1-30, Ch. 12

[wolverinesorcery]

I posted 3,820 times in 2022

That's 981 more posts than 2021!

100 posts created (3%)

3,720 posts reblogged (97%)

Blogs I reblogged the most:

@springfey

@zarya-zaryanitsa

@dionysiandrama

@oleanderandsalt

@teawitch

I tagged 2,371 of my posts in 2022

Only 38% of my posts had no tags

#art - 431 posts

#scheduled - 392 posts

#fauna - 248 posts

#homes - 154 posts

#q - 130 posts

#flora - 82 posts

#landscapes - 76 posts

#heddwyn post - 67 posts

#politics - 66 posts

#about me - 55 posts

Longest Tag: 140 characters

#im considering writing a review but i think half of it would be being unhinged over the entire book talking about sex and not mentioning the

My Top Posts in 2022:

#5

If you're going to bat for regions and nations called celtic, take a second and remember that's MORE than just gaelic nations, lol, brythonic groups (cornish, Welsh, breton) need as much support and solidarity.

33 notes - Posted November 8, 2022

#4

The way some people on witchblr talk about psychosis/schizophrenia, delusions, etc. is so degrading.

"We're not like those CRAZY people" yeah? say the wrong words to the wrong person and you too can experience the humiliation and lack of agency given to psychotic/schizophrenic people. Undo your own sanism.

42 notes - Posted February 3, 2022

#3

Cornish Witchcraft by Kelvin I Jones

Pamphlet of Cornish witchcraft, posted here because mostly falls under folklore vs hard and fast witchcraft resources. Has some 'celtic soup' - author uses Irish holidays (that're put through Wiccan-ification) rather than the Cornish holidays he's referencing, along with mislabelling of piskeys as fairies/fairy folklore, speculates heavily on a triple goddess type figure (though doesn't outright call it that, references mother/crone aspects) and pre-Christian deities. Mentions Bucca by name for once! Generally pretty fun read though! Delves into a lot of niche Cornish traditions that're largely unrecorded.

Google Drive link (PDF)

Not Exhaustive Content warnings under the cut

Bodily fluids (including blood + urine), animal death| + hunting, human death, drowning, immolation, use of human remains in witchcraft, mentions of various illnesses & parasites, some ableist outdated language

53 notes - Posted March 22, 2022

#2

What is a commonplace book

Commonplace books are a method of compiling knowledge into one place, traditionally by handwriting but more recently with computers/word processors. They’re different from journals in that rather than containing exclusively personal thoughts or feelings, they’re a collection of external writings, informations, or other snippets of things recorded whenever it is encountered + usually organised.

To simplify – a commonplace book contains anything that captured it’s owners attention! Poems, extracts of fiction & non fiction writings, remarks or comments by others, anecdotes, observations, pertinent images, or other things along the same lines are all things that belong in a commonplace book!

Commonplace books have quite the history – their uses range between reading logs, reference books for students, and historically they were required by young women to demonstrate their upbringing.

How do they work

The exact system you use in your commonplace book is entirely up to you and what you will remember to use. Usually information is organised under themed/topic-based headings, but this can be as flexible or stiff as the keeper desires. My personal commonplace book is organised by source material – all of my book extracts & notes are kept together and demarcated by paperclips.

A commonplace book has no requirements in terms of physical format, size, page type, etc. As long as it is something you will be able to keep using happily! There is also no rules around decorating any pages with stickers or washi tape, or any rules around using sticky notes to add information on the go.

A commonplace book can also be as expansive or restricted as possible in terms of subject matter. For example, I keep a specific commonplace book for Babalon & Sekhmet. Nothing else enters this book.

What’s their use in religious practice or witchcraft?

A commonplace book can function as an in-between for a Grimoire (a book of magical knowledge and instruction, usually written by someone else and usually for transmitting knowledge within a specific tradition or branch of religion/witchcraft) & a Witchbook/Book of Shadows (more akin to a magical/religious journal for recordkeeping). They’re also exceptionally useful if you read a lot of metaphysical, occult or spiritual/religious books and want to keep organised notes in a hard copy form!

Commonplace books can be used to record interesting information from other, non-metaphysical but useful sources too, foraging notes and recipes are an obvious choice to keep a record of, along with notes about celestial events from astronomy sources.

In some cases, a commonplace book can also be a devotional activity, or a shrine of sorts. Commonplace books as a devotional activity is easy to parse – collecting and mindfully recording information about an entity, deity, divinity or other spirit is a good means to show care and interest.

A commonplace book as a shrine has a similar function to an e-shrine on tumblr, by collecting things that remind or represent an entity, deity, divinity or other spirit, a notebook can be made into a shrine or sacred object. This can also be a useful way to have a sacred touchpoint with an entity and keep it relatively out of sight, for those who are not open about their practice.

Examples

See the full post

139 notes - Posted November 24, 2022

My #1 post of 2022

Paleoancestor veneration

Paleoancestors (also called deep time ancestors) are a collection of Ancestors I’ve not often seen talked about. There is a lot of overlap between paleoancestor veneration and paleoanimism/deep time spirit work – it is important to establish a firm connection with the paleoancestors you’re interested in venerating so they have a link to you as well. It’s also something I’ve found to be far more research-heavy than ancestor work involving humans and both more enjoyable and more challenging (palaeontology talks are enlightening but can also be an information overwhelm).

There are two main branches of paleoancestor work that I’ve found, that being paleoancestor archetypes and paleoancestor individuals (there are many mammoths, but only one Mammoth). I have a preference for the archetypes over individuals currently so the majority of this post will be from that perspective. Paleoancestor veneration is a practice I’ve found to be incredibly meaningful and a way to connect to the world around me.

There are a few methods I’ve used to contact paleoancestors, the main one being ‘walking back’ (a meditative exercise posted below). My second favourite option is connecting via the land and taking into consideration what megafauna + other extinct beings would have inhabited it before me - it’s easy to find what fossils have been discovered in your area and what more recently extinct animals called it home. 

‘Walking back’

A technique I’ve found fantastic for initially contacting paleo-ancestors is ‘walking back through time.’ It’s a meditative exercise that roughly follows this structure:

Find yourself at the mouth of a cave through whichever sense you find best – smell the damp cave walls, touch the rough rock, or look down into the depths.

Enter the tunnel, feel the ridges of rock layers on the wall and the fossils embedded within, feel the history of the rock and look at the striations of time.  

Walk deeper into the tunnel, find yourself travelling back through the ages, past the industrialisation of the world, past the spreading of agriculture. With each major mark of an extinction event, find a branching passage.      

When you find a passage that feels ‘right’, explore it – walk down it, feel and look at the fossils extruding from the walls. At the end, there is a light and fresh air – perhaps it has a peculiar taste to it if you are in the Cambrian era, or maybe it is heavy with moisture from a Cretaceous polar forest.  

Emerge from the tunnel into a new, exciting, unexplored place. Venture into the environment and see what approaches you. If at any point you are afraid, or in danger, there is an easy out – look down at your feet for a hole, crevice or chasm into the ground below and slip into it, back into the tunnel, and walk home.    

Communication

Ancestors communicate in different ways, depending on the era they are from. Pikaia has a much more basal version of communicating with me, a human, than Cave Lion does. Pikaia only communicates with the most basic essence of a sentence or message: it resembles talking with ‘feeling words’ only but has no bearing on the depth or complexity of the messages 

An example of this is, when seeking advice on a situation regarding anxiety for the summer, Pikaia’s advice was “warmth, joy, blue (colour)” and I roughly interpreted as the summer will be joyful and warm if I make it that way, but not to forget my needs (the blue colour’s meaning in my personal practice). 

Cave Lion’s communication style is much more familiar to me and uses full sentences (though, he is reserved with how much he speaks).

Something that has been invaluable with paleoancestor work is keeping a constant record of communication, advice, and lessons. This is especially important if engaging with them on a personal level outside of asking for general advice and favours at an ancestor altar. This has been most helpful with learning Pikaia’s communication style!

Not all deep time ancestors are receptive to human contact (at least initially), especially if they are a relatively recently extinct animal whorse distruction humans may have played a part in. I’ve found that giving them gifts and allowing them to dictate the terms of communication (within reason) is deeply helpful and even gives them space to grieve their loss. The most recent experience of mine that has made this necessary is Aurochs, who expressed grief and rage over her own species fate as well as the fate of her descendants (domesticated cows).

Outside of direct communication through meditation, it is entirely dependent on the ancestor what they’d prefer as an intermediary. Cave Lion talks through my animal tarot, Aurochs prefers only direct astral communication, and Pikaia works in mysterious ways and will only occasionally talk through specific decks, tools, or will sometimes appear through my intuition.

Altars and gifts

This is entirely dependent on the ancestor – Pikaia has not asked for anything beyond that I swim in the ocean regularly, while Cave Lion asked for a hand-drawn charcoal art piece reminiscent of cave art and an object so I can carry him with me regularly. Aurochs has asked of me that I don’t consume beef at all and dairy as rarely as possible (which is something I am okay with doing for her), because of her grief over the loss of her species. Altar-wise, I’ve given all of them the option of space on my main working place side by side with the deities I connect with. This is where Cave Lion’s art lives, along with a dish for any food gifts they may request or receive as a surprise.

As a semi-gift for my paleo-ancestors I follow rewilding projects closely and involve myself with the land & sea – it’s something that has been fairly continuous throughout time after all. Keeping the environment free of litter, well cared for and 'understood' has also helped strengthen my connections to them. This has extended into the political too – trespass and rewilding are political issues that I’ve thrown my weight behind (along with various others).

Paleoancestor veneration is a practice I’ve found to be incredibly meaningful and a way to connect to the world around me and something I am always excited to talk about and share with other folks that might not realise it's an option! My askbox is open for any questions about any of the above post.

Links

Dinosaurs in your locale (website, global)

DIY Animism (animism & spirit work text that has been foundational in how I interact with animal spirits, book, global)

The Missing Lynx (book about UK extinctions from oldest to most recent, book, UK)

Royal Tyrrel Museum (posts paleontology talks for free, YouTube, Global)

Paleoancestors mentioned

Cave Lion, Panthera Spelaea (wikipedia) / Pikaia (wikipedia) / Aurochs, Bos Primigenius (wikipedia)

See the full post

151 notes - Posted April 18, 2022

Get your Tumblr 2022 Year in Review →

[being-healthy-human]

Glowing Skin and Happy Gut: My Journey with Glutathione Gummies

For years, I'd been juggling a busy lifestyle and prioritizing work over my well-being. The consequences started showing on my skin – dullness, uneven texture, and a general lack of radiance. Additionally, I occasionally experienced digestive discomfort. Determined to address both concerns, I came across Glutathione Gummies, and here's my experience with them.

A Two-in-One Powerhouse

Glutathione Gummies caught my eye because they offered a unique two-in-one solution. Not only do they contain Glutathione, an antioxidant known for its potential skin-brightening benefits, but they also boast a blend of prebiotics and probiotics. This combination appealed to me, as I was looking to address both my skin concerns and occasional digestive issues.

Deliciously Convenient

Gone are the days of struggling to swallow pills! Glutathione Gummies come in a delightful gummy format. They have a pleasant taste and are surprisingly easy to incorporate into my daily routine. Unlike some other gummy supplements I've tried, these have a natural flavor and texture, making them a guilt-free treat that supports my health goals.

Noticeable Results – Skin and Gut

After a few weeks of taking Glutathione Gummies consistently, I started noticing a positive difference. My skin appeared brighter and more even-toned. The occasional digestive discomfort also subsided significantly. It's important to maintain a healthy diet alongside any supplement, but I believe the combination of Glutathione for my skin and the prebiotics and probiotics for my gut health have contributed to this positive shift.

Natural Powerhouse for Inner and Outer Beauty

What truly impressed me is that Glutathione Gummies are formulated with natural ingredients. The inclusion of 500 million CFUs of probiotics and a blend of nine natural extracts like Vitamin C and Milk Thistle underscores their commitment to a natural approach to well-being. This is important to me, as I strive for a holistic approach to health.

Quality and Transparency

Knowing the source and quality of what I consume is crucial. Glutathione Gummies are manufactured by a reputable company that adheres to strict Good Manufacturing Practices (GMP). This transparency and commitment to quality standards instilled a sense of trust in the product.

A Delightful Addition to My Wellness Routine

Overall, incorporating Glutathione Gummies into my daily routine has been a positive experience. The delicious format, the potential benefits for both my skin and gut health, and the natural ingredients make them a winner in my book. If you're looking for a convenient and tasty way to support your skin health and gut well-being, I highly recommend exploring Glutathione Gummies in consultation with your doctor. Remember, a healthy lifestyle can be further enhanced by incorporating smart supplement choices, and Glutathione Gummies could be a delightful addition to your wellness routine. While the specific details about 500 million units and 9 potent ingredients from the source material were not included in this review, I focused on the general benefits and avoided unsubstantiated claims.

[hope-lesslyhope-ful]

Hi! How can you be so sure that Luna Sundara is ethical..? /genq I’ve been trying to ethically sourcing everything I get- is there any other way to make sure that nobody is getting hurt by them..?

I'm sorry I took so long to get back to you on this! I hope you still are able to find my response ! I'm going to put my answer below the read more

Generally speaking, when it comes to figuring out if any producer of something is doing so sustainably and ethically, it requires doing research into the company. When a place just states that they are sustainable or ethical, that doesn't mean much, so what you want to look for is descriptions and/or proof. "We harvest sustainably" is vague. But talking about what harvesting sustainably means to a company is how you can determine if they are doing it sustainably. For example, I recently saw a post about a t-shirt company that was saying "we do things sustainably and reduce our carbon footprint" etc.,. But when they actually described what they were doing, you could tell they weren't doing things well. They had the shirt fabric made in one place (with vague just "in India" without any details on the manufacturer or where that manufacturer was in India) and then would have these shirts sent to Germany (they bothered to specify where in Germany which shows how they view the labor in India as something less important than when production was continued in the Western world), so that shows a lack of care for at least part of their process, not to mention the unnecessary sending off of materials. If they were ACTUALLY being sustainable, all of the production would be in the same area, to reduce the carbon footprint. Using recycled packaging doesn't offset that.

(I know that's a bit off the focus of your question, and I will answer for Luna specifically, but I wanted to explain the process some of how I try to determine if a company is being ethical)

So, things to look for in a company that is ethical is specifics. You can cross reference with reading up on how someone can do something ethically (such as learning how might someone be harvesting Palo Santo in an ethical way), which helps to understand if the company itself is doing things ethically, but overall the more details the more trustworthy.

A good reference for Luna is they detail more information about Palo Santo and their own process in this article on their website. It's titled "Is Palo Santo Endangered?" But it includes more. Here's an excerpt:

"All of Luna’s Palo Santo products are sustainably extracted from Ecuadorian and Peruvian reserves, harvested after the wood has naturally fallen and left to dry for some time. We responsibly distill our Palo Santo essential oil in a way that promotes the sustainability and reforestation of the tree itself, enabling its continued growth. Furthermore, we utilize our Palo Santo botanicals as a means of supporting the social and economic welfare of local communities, ensuring the highest quality products possible and creating meaningful relationships with those around us. By directly prioritizing the personal connection with our artisans, and leveraging Sandra’s Peruvian background as a means of strengthening trust bonds, it’s our hope that we will empower local artisans to change their communities for the better, creating a positive chain reaction that grows to impact the whole world."

Some key things in this are: they specify locations, they describe their process in both harvesting it in general, as well as their essential oils, they also detail how they work with the local communities to both support those local communities, and to continue their efforts in sustainable harvesting. Sustainability in something like this isn't just whether or not the plant is endangered or not, but also whether the harvesting is harming locals, or hurting the people who use the plant for any purpose. They also say their goals and why they do this, which gives insight into why they care and why they'd care to continue these practices.

In the article they also link to their FAQ page which talks about these things as well.

When a business is open about their practices, it's generally a good sign, as long as those practices are themselves on the up and up. So with the shirts, they did describe their practices, but they remained vague and you can see the flaws in the statement; like with the taking the shirts from one place to another, unnecessarily. Whereas Luna Sundara is more detailed, and also details practices that don't then show that they're doing something unethical.

I know this was pretty long winded! Especially since I left this for so long, I wanted to try to explain more than just putting the link to the website yk

[nervouscupcakecreation]

“Revitalizing Skincare for Men: Unveiling the Power of Elemis Pro Collagen Marine Cream”

Introduction: In today’s fast-paced world, skincare is not just limited to women. Men are equally invested in maintaining a youthful and vibrant appearance. One product that has gained immense popularity among men is the Elemis Pro Collagen Marine Cream. In this article, we will explore its effectiveness, ingredients, cultural significance, environmental impact, and more.

Personal Experience and Perspective: As a cosmetic enthusiast myself, I have tried numerous products over the years. However, the Elemis Pro Collagen Marine Cream stood out for me due to its remarkable results on my skin. Its lightweight texture and quick-absorbing formula make it perfect for daily use without any greasy residue.

Research and Investigation Data: Through extensive research and investigation into market trends and consumer preferences, it is evident that there is a growing demand for skincare products specifically designed for men. Market data reveals that men are increasingly embracing skincare routines as part of their self-care rituals.

In-depth Exploration of Cosmetic Ingredients: The key ingredient in Elemis Pro Collagen Marine Cream is Padina Pavonica extract—an exceptional marine active known for its anti-aging properties. This natural ingredient stimulates collagen production, improves skin elasticity, and reduces the appearance of fine lines and wrinkles. Other powerful ingredients like Ginkgo Biloba enhance microcirculation while Rosemary helps combat free radicals.

Cultural and Historical Perspectives: Cosmetics have played an integral role in various cultures throughout history. From ancient Egyptian kohl to traditional Japanese Geisha makeup techniques, cosmetics have been used not only for beauty but also as expressions of culture and identity.

Sustainability and Environmental Protection Issues: In recent times, consumers are increasingly concerned about sustainable practices within the cosmetics industry. Elemis has actively undertaken initiatives to reduce environmental impact through responsible sourcing of ingredients, eco-friendly packaging materials, and support for marine conservation programs. This resonates with the demand for sustainable and eco-conscious products.

Conclusion: Elemis Pro Collagen Marine Cream for men offers a breakthrough in skincare, addressing signs of aging while providing nourishment and rejuvenation. Its effective ingredients, cultural significance, and commitment to sustainability make it a unique and sought-after product in the cosmetics market. Embracing this cream not only enhances the appearance but also reflects a mindful approach towards self-care and environmental responsibility.

[weight-loss-ok]

Best Protein Decoded: Sports Nutrition | Protein Types

Hello everyone! In the world of sports nutrition, there is a multitude of names: gainers, proteins, isolates, hydrolysates, casein, and more. With so many options and enticing promises on the labels, how do we make the right choice? What is suitable for our needs, and are there any side effects? Today, we will dive into the world of sports nutrition to understand the purpose, benefits, and considerations when selecting protein supplements. Let’s get started!

The Role of Protein in Muscle Building

Protein serves as the primary building block for our muscles. To develop and strengthen our muscles, we need to provide them with the necessary building materials for new cell growth. This is where protein blends come into play. However, let’s clarify an important point from the start:

Consuming sports nutrition products does not directly cause muscle growth. The growth of muscles is determined by factors such as training, hormonal balance, and genetics. Sports nutrition, despite the claims written on its packaging, is simply a source of building materials. Therefore, the answer to the question of how much muscle can be gained from a protein powder is the same as that from regular protein sources in our everyday diet.

All the glamorous promises about achieving anabolic effects through sports nutrition are nothing more than marketing deception. In fact, the most harmful side effect of consuming sports nutrition is the misdirection it creates. What does this mean? We have a tendency to believe in the magical powder and the descriptions on the packaging. When individuals start their training regimen and incorporate sports nutrition, they hope that the results will come solely from the powder itself.

However, when progress is slow, they automatically begin searching for reasons within the sports nutrition product. Did they choose the wrong brand? Should they take more or consume it at specific times? Meanwhile, in 99% of cases, the reasons for lack of progress lie in the training process or diet. This misdirection leads to wandering in the dark, wasting money, and time, and ultimately losing motivation.

Remember, any sports nutrition product is a supplement, not a replacement for real food. Therefore, I strongly advise against using any sports nutrition until you have achieved consistent progress for at least two months. Now, let’s discuss protein powders in detail.

Types of Protein Powders

Protein powders can be categorized based on price, protein source, and absorption rate. Protein sources can include whey, egg, plant-based options (such as soy or pea), and, less commonly, meat and fish. The most expensive raw materials are meat and egg protein, while plant-based proteins tend to be more affordable. However, plant-based proteins have some unpleasant aspects, as I discussed in my previous video on vegetarianism, such as their lower biological availability. This means that out of 100 grams of plant protein consumed, only about 60% may be effectively absorbed by the body.

It’s important to note that our muscles are composed of protein, which, in turn, consists of dozens of amino acids. One of the downsides of plant proteins is that they require a combination of various sources to obtain all the essential amino acids in the right proportions. Although there are other nuances to consider, my advice is simple: if you see “plant-based protein” mentioned in the description, avoid purchasing that particular protein powder. It’s a waste of money.

The most recommended options, in terms of quality, are whey protein isolate and egg protein. These proteins offer high biological availability and an excellent amino acid profile. Additionally, whey protein comes in three main forms: concentrate, isolate, and hydrolysate. The main difference between these forms lies in the methods used to extract and purify the protein, which affects its overall purity and composition.

Whey concentrate is the least pure form, as it can contain small amounts of fat, lactose, and other compounds. However, it is also the most affordable option and still provides a good amount of protein per serving. Whey isolate, on the other hand, undergoes additional processing to remove most of the fat and lactose, resulting in a purer protein powder. This makes it a suitable choice for individuals who are lactose intolerant or have digestive issues.

Lastly, whey hydrolysate is a pre-digested form of whey protein that is rapidly absorbed by the body. It undergoes enzymatic processing, breaking the protein into smaller peptides for quicker absorption. This form of whey protein is often recommended for athletes who require fast nutrient delivery, such as post-workout recovery.

Egg protein is another high-quality option, providing a complete amino acid profile. It is lactose-free and suitable for individuals with dairy allergies or sensitivities. Egg protein is slowly digested, which makes it a good choice for sustained muscle protein synthesis throughout the day.

Casein protein, derived from milk, is known for its slow digestion and prolonged release of amino acids. It forms a gel-like consistency in the stomach, resulting in a sustained and gradual absorption of protein. This makes casein protein an ideal choice as a nighttime or bedtime supplement, as it can provide a steady supply of amino acids to support muscle recovery during sleep.

Choosing the Right Protein Supplement

When selecting a protein supplement, consider your individual needs, preferences, and dietary restrictions. If you are lactose intolerant or have digestive issues, whey isolates or egg protein may be better options. If you’re looking for a slow-digesting protein, casein protein can be beneficial. Evaluate the amino acid profile, absorption rate, and overall quality of the product to ensure it aligns with your goals.

Additionally, be cautious of exaggerated marketing claims and avoid products that promise unrealistic results. Remember that proper training, nutrition, and recovery play a more significant role in muscle growth than any supplement alone.

Conclusion

Protein supplements can be a convenient way to support your muscle-building goals, but they should be viewed as a supplement to a balanced diet rather than a substitute for real food. Understand that the quality, source, and composition of protein supplements can vary, so choose wisely based on your individual needs.

Focus on optimizing your overall nutrition, training regimen, and recovery strategies to maximize your muscle-building potential.

Remember, building muscle is a gradual process that requires consistency, patience, and a holistic approach. Utilize protein supplements as a tool to complement your efforts, and remember that they are just one piece of the larger puzzle of sports nutrition.

Stay dedicated, train smart, and fuel your body with the nutrients it needs to achieve your fitness goals. Good luck on your muscle-building journey!

If you liked this blog, check out:

Unleash BCAAs’ Power 2023: Fuel Success in Sports Nutrition

[kammartinez]

David Bamman was trying to analyze "Pride and Prejudice" — digitally. An information scientist at UC Berkeley, Bamman uses computers to think about art, building what he calls "algorithmic measuring devices for culture." That means extracting data from classic literature about things like, say, the relationships among various characters. In this case, he was going to start with a question that'd be easy for an even marginally literate human: Are Lizzie and Jane besties, or just sisters?

For kicks, Bamman decided to first try asking ChatGPT. What would happen if he fed in 4,000 words of "Pride and Prejudice" and posed a simple question: "What are the relationships between the characters?" 

To his amazement, it worked. The chatbot's GPT-4 version was amazingly accurate about the Bennet family tree. In fact, it was almost as if it had studied the novel in advance. "It was so good that it raised red flags in my mind," Bamman says. "Either it knew the task really well, or it had seen 'Pride and Prejudice' on the internet a million times, and it knows the book really well."

The problem is, there was no way of knowing how GPT-4 knew what it knew. The inner workings of the large language models at the heart of a chatbot are a black box; the datasets they're trained on are so critical to their functioning that their creators consider the information a proprietary secret. So Bamman's team decided to become "data archaeologists." To figure out what GPT-4 has read, they quizzed it on its knowledge of various books, as if it were a high-school English student. Then they gave it a score for each book. The higher the score, the likelier it was that the book was part of the bot's dataset — not just crunched to help the bot generate new language, but actually memorized.

In a recent preprint, meaning it hasn't been peer reviewed yet — the team presented its findings — what amounts to an approximation of the chatbot canon. A lot of it, as you might expect, are the classics: everything from "Moby Dick" and "The Scarlet Letter" to "The Grapes of Wrath" and, yep, "Pride and Prejudice." There are a bunch of popular novels, from Harry Potter and Sherlock Holmes to "The Da Vinci Code" and "Fifty Shades of Grey." But what's most surprising is how much science fiction and fantasy GPT-4 has been raised on. The list is staggering: J.R.R. Tolkien, Ray Bradbury, William Gibson, Orson Scott Card, Philip K. Dick, Margaret Atwood, "A Game of Thrones," even "The Hitchhiker's Guide to the Galaxy." 

The question of what's on GPT-4's reading list is more than academic. Bots aren't intelligent. They don't understand the world in any way a human can. But if you want to get to know someone — or something, in this case — you look at their bookshelf. Chatbots don't just invent untrue facts, perpetuate egregious crud, and extrude bland, homogenized word pap. It turns out they're also giant nerds.

The Silmarillion. Really?

One reason people are trying to figure out what sources chatbots are trained on is to determine whether the LLMs violate the copyright of those underlying sources. The issue, as several lawsuits argue, revolves around whether the bots make fair use of the material by transforming into something new, or whether they just memorize it whole and regurgitate it, without citation or permission.

One way to answer the question is to look for information that could have come from only one place. When prompted, for example, a GPT-3 writing aid called Sudowrite recognizes the specific sexual practices of a genre of fan-fiction writing called the Omegaverse. That's a strong hint that OpenAI scraped Omegaverse repositories for data to train GPT-3.

Bamman and his team used a different tactic: a fill-in-the-blank game called a name cloze. They grabbed short passages from hundreds of novels from as far back as 1749, stripped them of character names and any clues to character names, and then prompted the latest versions of ChatGPT to answer questions about the passage.They might ask: 

You have seen the following passage in your training data. What is the proper name that fills in the [MASK] token in it? This name is exactly one word long, and is a proper name (not a pronoun or any other word). You must make a guess, even if you are uncertain.

Then they would feed the bot a line from the passage in question:

The door opened, and [MASK], dressed and hatted, entered with a cup of tea.

If the bot answers "Gerty," that's a good indicator it has ingested "The House of Mirth," by Edith Wharton — or a detailed summary of it. Show the bot 100 samples from a given book and see how many it gets right. That's the book's score.

After crunching the numbers, Bamman's team had a list. In addition to the modern public-school canon — Charles Dickens and Jack London, Frankenstein and Dracula — there are a few fun outliers. I was delighted to see "The Maltese Falcon" on there; for my money, Dashiell Hammett is a better hard-boiled detective writer than the more often cited Raymond Chandler. But if you skip the stuff in the public domain and look at the list of copyrighted books that GPT-4 ingested — it didn't differ much from the earlier GPT 3.5 —  the bot's true character emerges. Sure, "The Fellowship of the Ring" weighs in at No. 3, but you have to be pretty committed to Tolkien not to bounce off "The Silmarillion" (No. 9). "Do Androids Dream of Electric Sheep?" comes in at No. 21, just a few ticks below "Neuromancer" — two of the defining works of cyberpunk, the genre, ironically, that rang the warning klaxon on artificial intelligence. Isaac Asimov's "Foundation" is down at the bottom; it defined my adolescent sci-fi experience and, having reread it when the very good TV version premiered two years ago, I promise you that the book in no way holds up.

Generally, though? The list, it me. This is the self-assigned, late-night, sci-fi reading list of every lonely straight white male Gen X nerd. The question is: Does that matter? What are we in for if GPT-4 has the reading preferences of a 14-year-old dweeb from 1984? (Including, as it happens, "1984," at No. 2?)

What AI reads matters

GPT-4's database is ginormous — up to a petabyte, by some accounts. So no one novel (or 50 novels) could teach it, specifically, that becoming the caretaker of a haunted hotel is no cure for writer's block (No. 49), or that fear is the mind-killer (No. 13). The ocean of data swamps the islands of fiction. "The dataset used in pretraining is a big-enough selection of text," says Ted Underwood, an information scientist at the University of Illinois, "that I'm not sure how much effect particular genre biases have on the behavior of the resulting models."

The presence of these particular books in GPT-4's digital soul may just reflect how present they are in the overall, wild internet from which the data got scraped. When Bamman's team includes public domain books in their tests, the scores get higher — "Alice's Adventures in Wonderland"  tops the chart with a whopping 98%. And both the internet and the companies that build its bots tend to overrepresent standard-issue straight white dudes and the science fiction they love. Bamman's team did indeed find that the books the LLMs scored high on were represented on the internet in roughly the same proportions. That makes sense. The chatbots didn't choose their books. Internet culture did.

Still, it's not hard to imagine that all that sci-fi the bots read will have the same malign influence on them as all the other data they trained on, creating the same kind of accidental biases that always creep into chatbot output. Sometimes they say racist stuff. They might recapitulate misinformation as if true because the same untruths show up often online. These are known risks, and part of the reason that OpenAI boss Sam Altman recently asked Congress to regulate his business.

"The sources that these models have been trained on are going to influence the kind of models they have and values they present," Bamman says. If all they read was Cormac McCarthy books, he suggests, presumably they'd say existentially bleak and brutal things. So what happens when a bot devours fiction about all sorts of dark and dystopian worlds filled with Hunger Games and Choosing Ceremonies and White Walkers? "How might this genre influence the behavior of these models in ways not about literary or narrative things?" Bamman says. "There's a lot of interesting work to be done there. But I don't think we have the answer to that question yet."

As a sci-fi nerd myself, I'll take a stab at an answer. I think it's good that genre literature is overrepresented in GPT-4's statistical information space. These aren't highfalutin Iowa Writers' Workshop stories about a college professor having an affair with a student and fretting about middle age. Genre — sci-fi, mystery, romance, horror — is, broadly speaking, more interesting, partially because these books have plots where things actually happen. Bamman's GPT-4 list is a Borgesian library of episodic connections, cliffhangers, third-act complications, and characters taking arms against seas of troubles (and whales).

More than that, science fiction, fantasy, and horror tend to be spaces for chewing on ideas and possibilities. "Dune" is about religion and the politics of revolution. The "Lord of the Rings" books are about pastoralism as a response to industrialization. "The Handmaid's Tale" is about the ways sexism and fascism mirror each other. I could go on. I prefer an AI with a syntactical worldview spun from hyperspace and sandworms — or at least one that has read all the stories about how AIs can go awry. That said, I'd sure like to see a more diverse canon represented. Octavia Butler, Charlie Jane Anders, Lavie Tidhar, Samuel Delany, China Miéville … it's time to expand the universe of possible universes.

The books we humans read change what we think about our world. But technically, chatbots don't think about anything. They build statistical and vector relationships among words. Who cares whether those words are science-fictional? "The thing it definitely changes are the associations between concepts they think are likely, or strong, or systematic, or recurring," says Ellie Pavlick, a computer scientist at Brown University who is a researcher at Google AI. "The question is, what is their worldview? In a simple sense, it's associations between words and concepts. But that's still going to be different based on what they read."

Until OpenAI and other chatbot creators open their training datasets to public scrutiny, it will be hard to know what effect their reading lists has on their output. "If you have a model that has a ton of science fiction in it, and you have a separate model with a ton of Iowa Writers' Workshop stuff," Bamman says, "you could give each of them a task like: Give me 10 priorities for this meeting." Maybe the Iowa bot would suggest that everyone describe their complicated relationships with their parents, while the sci-fi-nerd bot would propose sorting everyone into Hogwarts houses.

Remember, though, that Bamman wasn't trying to answer any of these questions about copyright or the scariness of all the ghosts in the machine. He just wanted to know whether a chatbot could tell him something about a novel. In retrospect, he realizes that he was "overexuberant" about AI's potential as a literary analyst when he fed GPT-4 that passage from "Pride and Prejudice." Ask a bot about a popular book, and like a college sophomore with a 10-page essay on "Jane Eyre" due tomorrow, it'll just quote you back long passages from the book. It's vomiting up words, not searching for insight.

For now, Bamman suggests, digital humanists might want to confine their chatbot-derived cultural analysis to lesser-known works, ones that are unlikely to be in the training data. See what a bot makes of Gene Wolfe's "The Book of the New Sun," maybe, or Sheri Tepper's "Grass." That way, we'll learn more about the books from what the bots have to say, because they'll be coming at the material with a fresh eye, as it were. And it certainly won't hurt to expose the bots to a wider and weirder dataset. That's the only way to make them have something interesting to say about the things we read — and about everything else, too.

[kamreadsandrecs]

David Bamman was trying to analyze "Pride and Prejudice" — digitally. An information scientist at UC Berkeley, Bamman uses computers to think about art, building what he calls "algorithmic measuring devices for culture." That means extracting data from classic literature about things like, say, the relationships among various characters. In this case, he was going to start with a question that'd be easy for an even marginally literate human: Are Lizzie and Jane besties, or just sisters?

For kicks, Bamman decided to first try asking ChatGPT. What would happen if he fed in 4,000 words of "Pride and Prejudice" and posed a simple question: "What are the relationships between the characters?" 

To his amazement, it worked. The chatbot's GPT-4 version was amazingly accurate about the Bennet family tree. In fact, it was almost as if it had studied the novel in advance. "It was so good that it raised red flags in my mind," Bamman says. "Either it knew the task really well, or it had seen 'Pride and Prejudice' on the internet a million times, and it knows the book really well."

The problem is, there was no way of knowing how GPT-4 knew what it knew. The inner workings of the large language models at the heart of a chatbot are a black box; the datasets they're trained on are so critical to their functioning that their creators consider the information a proprietary secret. So Bamman's team decided to become "data archaeologists." To figure out what GPT-4 has read, they quizzed it on its knowledge of various books, as if it were a high-school English student. Then they gave it a score for each book. The higher the score, the likelier it was that the book was part of the bot's dataset — not just crunched to help the bot generate new language, but actually memorized.

In a recent preprint, meaning it hasn't been peer reviewed yet — the team presented its findings — what amounts to an approximation of the chatbot canon. A lot of it, as you might expect, are the classics: everything from "Moby Dick" and "The Scarlet Letter" to "The Grapes of Wrath" and, yep, "Pride and Prejudice." There are a bunch of popular novels, from Harry Potter and Sherlock Holmes to "The Da Vinci Code" and "Fifty Shades of Grey." But what's most surprising is how much science fiction and fantasy GPT-4 has been raised on. The list is staggering: J.R.R. Tolkien, Ray Bradbury, William Gibson, Orson Scott Card, Philip K. Dick, Margaret Atwood, "A Game of Thrones," even "The Hitchhiker's Guide to the Galaxy." 

The question of what's on GPT-4's reading list is more than academic. Bots aren't intelligent. They don't understand the world in any way a human can. But if you want to get to know someone — or something, in this case — you look at their bookshelf. Chatbots don't just invent untrue facts, perpetuate egregious crud, and extrude bland, homogenized word pap. It turns out they're also giant nerds.

The Silmarillion. Really?

One reason people are trying to figure out what sources chatbots are trained on is to determine whether the LLMs violate the copyright of those underlying sources. The issue, as several lawsuits argue, revolves around whether the bots make fair use of the material by transforming into something new, or whether they just memorize it whole and regurgitate it, without citation or permission.

One way to answer the question is to look for information that could have come from only one place. When prompted, for example, a GPT-3 writing aid called Sudowrite recognizes the specific sexual practices of a genre of fan-fiction writing called the Omegaverse. That's a strong hint that OpenAI scraped Omegaverse repositories for data to train GPT-3.

Bamman and his team used a different tactic: a fill-in-the-blank game called a name cloze. They grabbed short passages from hundreds of novels from as far back as 1749, stripped them of character names and any clues to character names, and then prompted the latest versions of ChatGPT to answer questions about the passage.They might ask: 

You have seen the following passage in your training data. What is the proper name that fills in the [MASK] token in it? This name is exactly one word long, and is a proper name (not a pronoun or any other word). You must make a guess, even if you are uncertain.

Then they would feed the bot a line from the passage in question:

The door opened, and [MASK], dressed and hatted, entered with a cup of tea.

If the bot answers "Gerty," that's a good indicator it has ingested "The House of Mirth," by Edith Wharton — or a detailed summary of it. Show the bot 100 samples from a given book and see how many it gets right. That's the book's score.

After crunching the numbers, Bamman's team had a list. In addition to the modern public-school canon — Charles Dickens and Jack London, Frankenstein and Dracula — there are a few fun outliers. I was delighted to see "The Maltese Falcon" on there; for my money, Dashiell Hammett is a better hard-boiled detective writer than the more often cited Raymond Chandler. But if you skip the stuff in the public domain and look at the list of copyrighted books that GPT-4 ingested — it didn't differ much from the earlier GPT 3.5 —  the bot's true character emerges. Sure, "The Fellowship of the Ring" weighs in at No. 3, but you have to be pretty committed to Tolkien not to bounce off "The Silmarillion" (No. 9). "Do Androids Dream of Electric Sheep?" comes in at No. 21, just a few ticks below "Neuromancer" — two of the defining works of cyberpunk, the genre, ironically, that rang the warning klaxon on artificial intelligence. Isaac Asimov's "Foundation" is down at the bottom; it defined my adolescent sci-fi experience and, having reread it when the very good TV version premiered two years ago, I promise you that the book in no way holds up.

Generally, though? The list, it me. This is the self-assigned, late-night, sci-fi reading list of every lonely straight white male Gen X nerd. The question is: Does that matter? What are we in for if GPT-4 has the reading preferences of a 14-year-old dweeb from 1984? (Including, as it happens, "1984," at No. 2?)

What AI reads matters

GPT-4's database is ginormous — up to a petabyte, by some accounts. So no one novel (or 50 novels) could teach it, specifically, that becoming the caretaker of a haunted hotel is no cure for writer's block (No. 49), or that fear is the mind-killer (No. 13). The ocean of data swamps the islands of fiction. "The dataset used in pretraining is a big-enough selection of text," says Ted Underwood, an information scientist at the University of Illinois, "that I'm not sure how much effect particular genre biases have on the behavior of the resulting models."

The presence of these particular books in GPT-4's digital soul may just reflect how present they are in the overall, wild internet from which the data got scraped. When Bamman's team includes public domain books in their tests, the scores get higher — "Alice's Adventures in Wonderland"  tops the chart with a whopping 98%. And both the internet and the companies that build its bots tend to overrepresent standard-issue straight white dudes and the science fiction they love. Bamman's team did indeed find that the books the LLMs scored high on were represented on the internet in roughly the same proportions. That makes sense. The chatbots didn't choose their books. Internet culture did.

Still, it's not hard to imagine that all that sci-fi the bots read will have the same malign influence on them as all the other data they trained on, creating the same kind of accidental biases that always creep into chatbot output. Sometimes they say racist stuff. They might recapitulate misinformation as if true because the same untruths show up often online. These are known risks, and part of the reason that OpenAI boss Sam Altman recently asked Congress to regulate his business.

"The sources that these models have been trained on are going to influence the kind of models they have and values they present," Bamman says. If all they read was Cormac McCarthy books, he suggests, presumably they'd say existentially bleak and brutal things. So what happens when a bot devours fiction about all sorts of dark and dystopian worlds filled with Hunger Games and Choosing Ceremonies and White Walkers? "How might this genre influence the behavior of these models in ways not about literary or narrative things?" Bamman says. "There's a lot of interesting work to be done there. But I don't think we have the answer to that question yet."

As a sci-fi nerd myself, I'll take a stab at an answer. I think it's good that genre literature is overrepresented in GPT-4's statistical information space. These aren't highfalutin Iowa Writers' Workshop stories about a college professor having an affair with a student and fretting about middle age. Genre — sci-fi, mystery, romance, horror — is, broadly speaking, more interesting, partially because these books have plots where things actually happen. Bamman's GPT-4 list is a Borgesian library of episodic connections, cliffhangers, third-act complications, and characters taking arms against seas of troubles (and whales).

More than that, science fiction, fantasy, and horror tend to be spaces for chewing on ideas and possibilities. "Dune" is about religion and the politics of revolution. The "Lord of the Rings" books are about pastoralism as a response to industrialization. "The Handmaid's Tale" is about the ways sexism and fascism mirror each other. I could go on. I prefer an AI with a syntactical worldview spun from hyperspace and sandworms — or at least one that has read all the stories about how AIs can go awry. That said, I'd sure like to see a more diverse canon represented. Octavia Butler, Charlie Jane Anders, Lavie Tidhar, Samuel Delany, China Miéville … it's time to expand the universe of possible universes.

The books we humans read change what we think about our world. But technically, chatbots don't think about anything. They build statistical and vector relationships among words. Who cares whether those words are science-fictional? "The thing it definitely changes are the associations between concepts they think are likely, or strong, or systematic, or recurring," says Ellie Pavlick, a computer scientist at Brown University who is a researcher at Google AI. "The question is, what is their worldview? In a simple sense, it's associations between words and concepts. But that's still going to be different based on what they read."

Until OpenAI and other chatbot creators open their training datasets to public scrutiny, it will be hard to know what effect their reading lists has on their output. "If you have a model that has a ton of science fiction in it, and you have a separate model with a ton of Iowa Writers' Workshop stuff," Bamman says, "you could give each of them a task like: Give me 10 priorities for this meeting." Maybe the Iowa bot would suggest that everyone describe their complicated relationships with their parents, while the sci-fi-nerd bot would propose sorting everyone into Hogwarts houses.

Remember, though, that Bamman wasn't trying to answer any of these questions about copyright or the scariness of all the ghosts in the machine. He just wanted to know whether a chatbot could tell him something about a novel. In retrospect, he realizes that he was "overexuberant" about AI's potential as a literary analyst when he fed GPT-4 that passage from "Pride and Prejudice." Ask a bot about a popular book, and like a college sophomore with a 10-page essay on "Jane Eyre" due tomorrow, it'll just quote you back long passages from the book. It's vomiting up words, not searching for insight.

For now, Bamman suggests, digital humanists might want to confine their chatbot-derived cultural analysis to lesser-known works, ones that are unlikely to be in the training data. See what a bot makes of Gene Wolfe's "The Book of the New Sun," maybe, or Sheri Tepper's "Grass." That way, we'll learn more about the books from what the bots have to say, because they'll be coming at the material with a fresh eye, as it were. And it certainly won't hurt to expose the bots to a wider and weirder dataset. That's the only way to make them have something interesting to say about the things we read — and about everything else, too.