top of page

Sun trine Pluto: Astrology, AI, and spirituality

The Devil’s Dictionary tells us that astrology is “[the] science of making the dupe see stars” (ASTROLOGY - The Devil’s Dictionary Dot-Com). Yet whether or not millions of astrology aficionados are truly dupes, Co–Star—an app that uses AI algorithms to generate daily horoscopes and natal charts—has led to an explosion of interest in astrology. In the United States, more than one-quarter of women aged 18-25 have downloaded the app, and the company is known for their cheeky social media posts that, when shared by users, act as free advertising for the app itself (Kokalitcheva). One such Instagram post from March 31, 2023, features white text on a black background with a large banner that reads “[desperately] wants you to notice.” For Leo, the response reads “[literally] anything,” while for Aries, it is “[that] they’re pointedly ignoring you” (“Co–Star on Instagram”). Understanding Co–Star primarily as a social media app that relies on an astrological lexicon, in this paper, I explore the app’s place in the world of astrology and its implications for spirituality. I examine the value of Co–Star’s reliance on natural language processing (NLP) to generate daily horoscopes, pithy content, and astrological charts. I ask a modified version of the original Turing question (Turing):[1] Can a machine approximate human spirituality through personalized astrological content? More importantly: does it matter?

Co–Star’s content matters to different users in different ways: for casual consumers with little knowledge beyond the zodiac sign associated with their birthdays, generating only a natal chart through AI—since the rest of the content is devoid of astrological vocabulary—may not matter very much. Many people instead engage with Co–Star as a meme, screenshotting content to send to friends to laugh at its occasional absurdity or contradictory horoscopes. For people with extensive knowledge of astrological phenomena, including but not limited to house systems, aspects, placements, and nodes, the randomness of the AI-generated content and its lack of basis on one’s charts—the original point of astrology—become major sticking points that decrease the app’s appeal beyond its value as a meme. Does AI then trivialize the role of astrology, poorly approximating or removing the human element of being in a community with others who flock to it for spiritual and emotional fulfillment? I argue that Co–Star may be useful for simple fun, watered down from its original intent to furnish individuals with genuine astrological insight; this shift indicates changing relationships with what we consider spiritual and our expectations of how “easy” spiritual communion should be.

How Co–Star Works

Co–Star emerged in 2017 as a spiritual pioneer, positing that astrology and AI could meld into an informative and communal experience for users and promising accessibility to those unfamiliar with planetary movements. The app attempts to make astrology intelligible to the general public and intends for its users to share the app’s content to other social media platforms, as well as with friends. To begin, one inputs the year, month, day, local time, and location of one’s birth. The app then generates one’s chart, comprising the placements of all relevant planets and the areas of life they govern—such as love, creativity, spirituality, or self. These planets include Mars, Venus, Mercury, Saturn, Jupiter, Uranus, and Pluto. The most important of the planets, however, are typically the sun, the moon, and the rising (or ascendant) sign. These comprise the “big three” of astrology and are the planets to which one refers in saying, for example, that one is a “Scorpio sun, a Taurus moon, and a Libra rising.” They are most frequently discussed on social media or in memes. The sun sign is the one with which people most commonly identify or which describes core personality traits, while the moon governs emotions, and the rising sign indicates one’s outwardly perceived personality.

One tab of the app, labeled “Chart,” explains what aspect of life each planet governs, as well as the significance of the sign that belongs to that planet and the sign’s house. For example, my sun is in Aries in the ninth house, therefore the app informs me that I am persistent and independent and that I “feel the need to distinguish [myself] from others through philosophy, faith, [and] education.” The part with which users are most likely to interact is the “Updates” section, which begins with a daily provocative headline, followed by three icons—a leaf, a fire, or a “no” emoji—which correspond to the relative ease with which one will interact with a specific dimension on that day. “Updates” is where daily horoscopes appear. For example, on April 12, 2022, my update informed me that I would experience “[power] in routine.” All these updates, including the paragraphs that contain words of wisdom, are assumed to be generated through NLP, though it is unclear how much text is created in this way. NLP is “the branch of artificial intelligence or AI […] concerned with giving computers the ability to understand text and spoken words in much the same way human beings can” (“What Is Natural Language Processing?”). NLP uses computational linguistics and other quantitative methods to model human language.

Co–Star’s own website copy, however, is where the true fun begins. Co–Star’s CEO, Banu Guler, explains that “today’s problems are bound up with the rise of technology” but also believes that “technology will be the antidote, by teaching us to speak about ourselves” (Smallwood). This intersection of AI and astrology justifies Co–Star’s existence. On the company’s About page, a text box reads:

Using NASA data interpreted by humans, Co–Star delivers personalized readings in real time. It starts by using the day, time, and place we were born to create an astronomical snapshot of the sky. This picture creates an in-depth personality analysis on character traits, behavioral tendencies, and emotional leanings: our astral chart. This chart is mapped against the current locations of the planets, signs, and aspects to show what our chart means today: our horoscope. …. Using artificial intelligence, Co–Star translates data into language we can read, understand, and share. (emphasis mine, Co–Star: Hyper-Personalized, Real-Time Horoscopes)

This vaguely corporate language leads to a perfectly valid question: to what extent are these astrological readings driven by AI? That is, where is the line between data and “human interpretation”? Co–Star is deliberately coy about revealing this information.[2] Christine Smallwood writes that “[the] app generates content by pulling and recombining phrases that have been coded to correspond to astronomical phenomena” (Smallwood) in order to generate daily readings. This process might mean “pulling” words or phrases associated with emotions or planetary positions. Ayoub Samadi writes that “the app uses proprietary AI technology to draw out the user’s birth chart from publicly accessible data gathered by NASA’s network,” but what that NASA data might be and how it is translated into astrological insights is unknown. It is possible that the data contains information regarding the positions of celestial bodies, which is then used by Co–Star’s algorithm in combination with phrases and sentences that are vaguely “astrological” to provide the app’s users with their daily wisdom. The parts of the app that have been affected by human interpretation are also unknown. Tellingly, all descriptions of their process, be it in-house or in the media, sound somewhat similar and do not shed light on the boundary between human language and NLP.

Despite Co–Star’s billing as an astrology app, in many places, the astrology simply disappears or is incorrect. One Reddit user in a Reddit “ask me anything” (AMA) discussion thread with Guler writes that “some details are foundationally wrong, like the ascendant being a ‘mask,’ which is not a commonly accepted idea present in modern or older branches of astrology I know of except pop astrology” (banu__guler).[3] The most notable part of Co–Star’s About page is the following sentence from its CEO: “This chart is mapped against the current locations of the planets, signs, and aspects to show what our chart means today: our horoscope” (Co–Star: Hyper-Personalized, Real-Time Horoscopes). While the app does generate a unique natal chart, the daily updates are often unmoored from it. On Sunday, April 10, 2022, the app told me that my “ability to empathize is high right now. … You know how valuable this is.”[4] One’s moon sign dictates one’s emotional state and personality. Is my moon in Aries then at least partially responsible for this insight about empathy? Or do other planetary movements and positions relative to each other also influence the update from the app? It is impossible to tell. The app does not explain the reason behind this increased empathy and generate further insights, nor does it show me how or why this interpretation is rooted in my chart, lowering the quality of this "astrological” information and the possibility for deeper spiritual engagement.

Horoscopes and other insights depend on information about relationships between the motions of various planets to yield daily or seasonal chart readings; this process is the cornerstone of astrology. Without examining how different aspects of one’s natal chart interact to create certain conditions in the present—for example, interpreting how two planets conjunct to each other affect emotions or decisions—Co–Star’s astrological insight is effectively rendered moot. The opacity of the data “interpreted by humans,” combined with the ostensible lack of relationship to one’s chart, actually renders Co–Star less of an astrology app and more one dealing in fancy buzzwords which capitalize on astrology’s ability to provide a shallow entry into spiritual communion with the self.[5] Indeed, Samadi notes that “Co–Star is fundamentally a data processing software first and an astrologist second.” In this light, the fuzziness of the boundary between the algorithmically-generated language and the “human interpretation” becomes not a quirk of the app but rather a deliberate choice that eliminates the traditionally human task of reading and interpreting astrological charts.

There are many ways to interact with the content of the app. One can tap the words on the landing screen to go directly to a “routine” update or go to the updates page to read all the day’s horoscope information all at once. Further interactions include tapping the thumbs up or thumbs down buttons at the bottom, and feedback that goes to the app’s developers to be used for potential updates or changes in the user interface (UI). This last action is particularly intriguing, as it is similar to receiving a chart reading from an experienced astrologer: they often say to take what resonates and leave the rest. The thumbs up or down thus acts as a proxy for what is useful or what is not—as well as what users might want to read in their daily updates. It thus asks us not to necessarily take “what speaks to us,” but to tell the algorithm whether we liked what the update had to say. This kind of feedback is more binary and certainly more prescriptive—over time, the algorithm learns what specific users may or may not want to hear. Such a feedback loop undermines the power of astrology to push individuals to reflect on themselves, their relationships, and their behaviour,[6] and reduces the generation of daily updates to a series of palatable missives. In relaying feedback within such a binary, we fine-tune the algorithm to ourselves until the output becomes not based in astrology—if it ever was, given that only the initial natal chart appears to discuss signs in-depth—but rather in the specific preferences of what an individual desires from Co–Star’s astrological “wisdom.”

What remains unclear is how decisive this user feedback is in Co–Star’s overall algorithm. It seems unlikely that a simple thumbs up or thumbs down is able to toggle the NLP-generated updates to such an extent that one’s daily notifications and “horoscope”[7] would change dramatically. The thumbs-up feature does, however, create the possibility that users will return daily to see the fluctuations in their “horoscopes” and anecdotally track these changes over time. Such changes provide the user with the impression that the “astrology” on which the app is based is “legitimate,” as even those mostly unfamiliar with astrology understand that the planets and stars move as often as daily. Ironically, however, the initial problem of being unable to separate the AI from “human interpretation” makes it difficult to gauge whether these fluctuations are rooted in astrological lore or the algorithm becoming smarter over time, with greater use, and with rapidly increasing numbers of users.

Why astrology?

If Co–Star does not provide an authentic astrological experience to its users, why is the app so popular? That is, what is so appealing about AI-driven astrology in the form of an app? First, everything is now an app, and this one, in particular, is appealing for its at-times bizarre daily updates that users can screenshot and send to friends. Inken Prohl writes: “[imagined] religious transcendencies are thus competing with digital transcendencies. Apps based on AI and digital offers pervade into areas such as mindfulness and counseling,” while Lee et al. note that “[algorithms’] complex, mathematically-based infrastructure is perceived to reveal objective, underlying truths rather than constructing new realities.” Quantifying spirituality lends it a powerful aura of authority in a culture that reveres technological solutions to abstract and open-ended questions such as those concerning spiritual life. Co–Star draws on NLP to generate easily-digested short paragraphs for app users; the combination of the pleasing monochromatic interface and the just-vague-enough daily updates have therapeutic effects that fit squarely within mainstream capitalist notions of “self-care” and “wellness” culture. Co–Star’s simplification of astrology can act as a painless and inoffensive dupe for mental healthcare: “It’s hard to understand the deep appeal of astrological practice without having or observing an individual chart reading, an experience whose closest analogue is therapy. But unlike therapy, … astrology promises to get to answers more quickly” (Smallwood). Co–Star’s place in the app ecosystem grants it the ostensible objectivity of technology, all the while filling the subjective emotional needs of its users through vague phrases and the cultivation of an ethereal yet somewhat distant speaker.

Theodor Adorno, writing about astrology in the 1950s, perhaps unknowingly foretold of Co–Star’s success; his disdain of the practice provides us with a better understanding of why Co–Star is so popular even if it is not rooted in nuanced understandings of astrology, nor always completely intelligible thanks to AI. (Indeed, NLP, plus “human interpretation,” is why the language is at times awkward in grammar and syntax.) Adorno writes that as fans of astrology “are gradually being transformed into things, the more they invest things with a human aura. … the libidinization of gadgets is indirectly narcissistic in as much as it feeds on the ego's control of nature—gadgets provide the subject with some memories of early feelings of omnipotence.” Users of Co–Star become the commodity—instead of the app itself—and ascribe human traits to the app’s language through this process of commodification. This phenomenon is borne out through the use of the app: apps can use different kinds of available data from users’ phones. Co–Star, through the use of NASA data and natal chart derivation, learns a few things about each user and can then tailor responses and output to something that we recognize as spiritual missives through prose.

Co–Star is an excellent case study for thinking about whether we are using machines or they are using us. Millions of users check the app frequently and read the daily updates to fulfill a wide variety of needs, be they emotional, spiritual, or simply entertainment. The app learns from our engagement. Indeed, Co–Star admits to tailoring the app’s appearance to people’s needs, in addition to making use of Google Analytics for “bug fixes”:

We choose which transits to surface on your home screen based on your past behavior within the app. For example, if you always look at your 6th/10th house or Saturn transits, we assume that’s because you’re a workaholic and show you more of those transits – and sometimes show you less because you should go outside more. (Frequently Asked Questions about Co-Star Astrology Society)

Such a feedback loop undermines the power of astrology to push individuals to reflect on themselves, their relationships, and their behaviour, and reduces the generation of daily updates to a series of palatable missives. In contrast to Adorno, Leona Nikolič observes that “media technologies such as smartphones and spiritual smartphone applications are not so much extensions of us as we are vessels for their viral obliteration of reality,” indicating that Adorno’s assessment of the fuzziness between human and object is perhaps both prescient and predictive. In commodifying the user when seeking their feedback and tracking their engagement, Co–Star’s AI algorithm can adapt to user “needs,” all while cloaked in a hazy lexicon that sounds somewhat astrological.

The lack of contextualization—and the calculated choice to relay to users that both AI and humans are interacting to produce “horoscopes”—makes Co–Star seem tailored to the user and gives the impression that the user is being acted upon by larger incomprehensible forces.[8] Lauren Oyler writes that “the main appeal is that [astrology] allows for ‘hyper-personalization’—the sense that one is special. … star signs were always a strategy for communicating a discrete set of qualities that are within your control.” Co–Star touts its focus on “hyper-personalization,” saying that “[most] horoscopes ask what month you were born. Co–Star asks what minute” (Co–Star: Hyper-Personalized, Real-Time Horoscopes). This description is somewhat misleading, as it is impossible to create a natal chart without this knowledge in the first place. Time, down to the minute, and location, are the very prerequisites for even making a chart. It seems as though Co–Star assumes that the app’s users may not know this. This obfuscation deliberately presents the app as a leader among its competitors while also making the user feel chosen or special, as though they have a particularly unique relationship with the app. While one’s sun, moon, and rising signs can appear in any of the 12 zodiac signs, thus permitting up to 1728 combinations, Co–Star is not alone in its ability to calculate these placements, despite its promise of hyper-specificity.

In reality, Co–Star elides the broader astrological context that is so crucial to the original practice in the first place, and the company indirectly admits to this in their explanation about what people can see when they open the app. The effect is one of human users interacting with an AI algorithm that cosplays as spiritual wisdom. Oyler, whose cultural criticism of internet culture is also applicable to Co–Star in this particular context, notes that “the internet is the ideal breeding ground for frivolous irrationality that incorporates the narcissistic character building of the personal brand.” Co–Star cultivates a sense of uniqueness in each user and crafts updates that only vaguely relate to one’s chart but which nevertheless speak to one’s desire for spiritual connection from a distance. Astrology is merely the vehicle for these goals.

Squaring the circle of AI, astrology, and spirituality

Co–Star’s popularity implies that if the algorithm sounds almost passably human, then the user’s interaction with the app and understanding of astrology will be frictionless to the point of indistinguishability between human and machine. Outside the scope of this paper, but no less interesting, is a systematic examination of Co–Star’s “horoscopes” over time; anecdotally, just a few years ago, they were riddled with grammar issues, sentence fragments, and incorrectly conjugated verbs. These issues still occur, but with less frequency. It seems that the algorithm, perhaps due to heavy user engagement and labour on the backend of the app, can now generate somewhat clearer sentences. But what does the lack of boundaries between NASA data, algorithms, and “human interpretation” mean for spiritual practices?

I suggest that, for casual users, Co–Star’s scant references to any astrological phenomena matter very little. To use the parlance of its main demographic, the app exudes impeccably esoteric “vibes,” or affects, of astrology that convey a sense that the app knows the user—an observation that is perhaps truer than many users know, given that the algorithm is in a constant state of iteration. Adorno writes that “[it] is as though astrology has to provide gratifications …, but is not allowed to interfere too obviously with the ‘normal’ functioning of the individual in reality.” That is, Co–Star’s vibes have to be carefully calibrated to repeatedly beckon users to their daily updates, a process achieved through milquetoast insights and quirky notifications. But for other users more invested in and dedicated to the spiritual experience that astrology provides, the app lacks the requisite information to interpret planetary and sign placements and their interactions. In sum: Co–Star speaks to the quick-fix bite-sized forms of “wellness” that attempt to ease contemporary life in the digital age.

In this milieu, regardless of the extent of users’ spirituality as pertaining to Co–Star, the mainstreaming of astrology via an app with a pleasing UI can have various implications for religious and spiritual life. Heidi Campbell writes that “employing digital, mobile technologies to support traditional religious practices is shaping many people’s understanding of religious praxis.” In other words, what religion is to an individual and how spirituality might be cultivated is shifting due to the foray of apps and increasingly sophisticated digital technologies. Jeremy Stolow posits that “the most fruitful studies often turn out to be those which proceed, not from the instrumentalist formula, ‘religion and media’ … but rather from the idea of ‘religion as media.’” Applying this idea to Co–Star, spirituality or feelings of community with others arise from the daily ritual of using the app, from seeing if one’s friends are doing well or if one’s friend’s chart yields any overlap with oneself. Changing conceptions of religion and spirituality manifest through the lens of digital media that is intended to be shared. Indeed, “many religious symbols and narratives become freed from their traditional structures and dogmas and so become tools for reconstructing spiritual meaning in daily life” (Campbell), suggesting that apps like Co–Star simultaneously change traditional understandings of what spiritual life can look like and reinforce the persona individuals seek within this sphere of their lives.

To return to the modified Turing question posed at the beginning of this essay: perhaps machine-generated text can produce some spiritual experiences for different types of users. The extent to which this is possible depends on the user’s initial desire to mediate spirituality through a social media app in the first place. Social media use is also highly ritualized, making an app the ideal place for repeated returns to astrology. Co–Star emphasizes that it offers a “personalized” experience, satisfying one’s need to be unique while also providing users with new ways to “add friends” on the app and connect with others in a rapidly-growing forum. Yet, while it turns astrology on its head and posits that AI can deliver some kind of spiritual guidance despite its basis in algorithms, the app ultimately may not offer anything new. People still crave human connection and spiritual fulfillment—it is simply that the means and methods have changed. Everything is data now—especially us. For surface-level engagement with various spiritual avenues, such as that of Co–Star’s take on astrology, perhaps this approach works. (“Surface-level” may not even be an entirely negative thing for the many people who now have easier access to at least some astrological knowledge.) Yet for those paying attention and who seek deeper levels of interaction with astrological principles and thus more from their spiritual experiences, the AI-generated content on Co–Star rings somewhat hollow beyond its shareability in a broader social media ecosystem. Beneath the personalization, the possibility to “connect” with friends on the app, the imprecise daily updates, and the cold yet appealing interface lies a deeper spiritual world that, it seems, algorithms and NLP not only obfuscate, but render inaccessible.



[1] His first question, though later revised, is: ‘Can machines think?’

[2] In a Reddit Ask Me Anything (AMA), a forum space where people can ask questions of any invitee, Guler says that the company’s content comes from “human staff astrologers [who] write everything with help from our homegrown AI.” She then clarifies that “homegrown AI” is “personalized content for every single person who uses the app,” and declines to comment on how, exactly, the algorithm generates that content.

[3] The “ascendant as mask” discusses the role of the ascendant or rising sign in one’s astrological chart, disputing the idea that the ascendant sign “masks” other traits or aspects of the self.

[4] This information is in Appendix 1, which I have withheld for anonymity.

[5] This is not to say that astrology is not spiritual or that it is only a shallow pursuit. For many people, astrology is indeed a profound spiritual practice, one that requires significant engagement over a long period of time and a deep understanding of both astronomical and astrological principles.

[6] Astrology is also important to many individuals for its ability to be more aware of the change in seasons, weather, and sky. Many people are more attuned to their energy levels, health, and natural surroundings; thus the notion of liking or disliking what the app presents makes little sense within the framework of a spiritual practice that encourages self-reflection and growth.

[7] I use this word in quotes because, as mentioned above, they are couched in non-astrological language and never reference planetary positions or interactions to contextualize their advice and insight.

[8] Again, ironically, astrology outside of Co–Star does address such relationships—between individuals, but also between the individual and the world around them—such that we can better understand how the planets and stars relate to us. Co–Star omits this communal aspect, though in recent months the app developers have initiated a chat forum at the bottom of every landing page to facilitate discussions among people with the same signs in the same planets. Introducing a chat feature complicates Co–Star’s place among followers of astrology and in larger discussions about its possible status as a platform.


Works Cited

Adorno, Theodor W. The Stars Down to Earth: The Los Angeles Times Astrology Column. p. 78.

ASTROLOGY - The Devil’s Dictionary Dot-Com. Accessed 18 Apr. 2022.

banu__guler. “Hi, I’m Banu, the Founder of Co–Star, AMA!” R/Astrology, 1 Dec. 2020,

Campbell, Heidi A. “Understanding the Relationship between Religion Online and Offline in a Networked Society.” Journal of the American Academy of Religion, vol. 80, no. 1, Mar. 2012, pp. 64–93,

Co – Star: Hyper-Personalized, Real-Time Horoscopes. Accessed 18 Apr. 2022.

“Co – Star on Instagram: ‘Virgos, Overworked and Underappreciated.’” Instagram, 31 Mar. 2023,

Frequently Asked Questions about Co—Star Astrology Society. Accessed 5 Mar. 2022.

Gillespie, Tarleton. Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions That Shape Social Media. Yale University Press, 2018,

Horst, Heather, et al. “Storing and Sharing: Everyday Relationships with Digital Material.” New Media & Society, vol. 23, no. 4, Apr. 2021, pp. 657–71. SAGE Journals,

Kokalitcheva, Kia. “Astrology App Co-Star Raises $15 Million in New Funding.” Axios, 14 Apr. 2021,

Lee, Joyce, et al. Exploring the “Magic” of Algorithmic Predictions with Technology-Mediated Tarot Card Readings. p. 43.

Light, Ben, et al. “The Walkthrough Method: An Approach to the Study of Apps.” New Media & Society, vol. 20, no. 3, Mar. 2018, pp. 881–900. SAGE Journals,

Nikolic, Leona. “Exploring Hyperreal Transcendence through Research-Creation: The Smartphone as Spiritual Interface between the Real and Virtual Selves.” The Journal of Media Art Study and Theory, vol. 2, no. 1, 2021, p. 11.

prohl, inken. “Material Religion and Artificial Intelligence.” Material Religion, vol. 17, no. 4, Aug. 2021, pp. 540–43. (Crossref),

Samadi, Ayoub. Tracking the Stars: Co-Star and the Creation of the Digital Celestial. 3 Oct. 2021,

Smallwood, Christine. “Astrology in the Age of Uncertainty.” The New Yorker, 21 Oct. 2019.,

Stolow, Jeremy. “Religion and/as Media.” Theory, Culture & Society, vol. 22, no. 4, 2005, pp. 119–45,

Turing, A. M. “I.—COMPUTING MACHINERY AND INTELLIGENCE.” Mind, vol. LIX, no. 236, Oct. 1950, pp. 433–60. (Crossref),

“What Is Natural Language Processing?” IBM, Accessed 26 Apr. 2023.

Date submitted: February 7, 2023

Date accepted: April 18, 2023

16 views0 comments

Recent Posts

See All

In response to “‘Uninterrupted Joy, Unrivalled Love’: Reading Paradise Lost Through Affective Prayer” by Shelby Haber. When I first encountered Paradise Lost as a young, bright-eyed undergraduate, I w

Psychosis, Self-Estrangement, and Estrangement from Reality in Robert Lowell’s Neurodiverse Poetry “And some go mystical, and some go mad.”[1]—A.M. Klein Abstract This paper illuminates the ways in wh

Throughout Jude the Obscure, Thomas Hardy pays a great deal of attention to clothing details, lingering on sartorial choices and changes characters make over the course of the narrative. From Arabella

bottom of page