pith. machine review for the scientific record. sign in

arxiv: 2604.15786 · v1 · submitted 2026-04-17 · 💻 cs.HC

Recognition: unknown

Filter Babel: The Challenge of Synthetic Media to Authenticity and Common Ground in AI-Mediated Communication

Advait Sarkar
Authors on Pith no claims yet

Pith reviewed 2026-05-10 08:09 UTC · model grok-4.3

classification 💻 cs.HC
keywords synthetic mediaAI-mediated communicationprivate experiencescommon groundauthenticityidentitythought experimentpersonalization
0
0 comments X

The pith

In a near future of fully private AI-generated experiences, individuals may develop languages intelligible to others only through AI translators.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents a thought experiment imagining a world where AI creates all content, media, and interactions uniquely for each person. In this setup, people could form distinct personal languages shaped by their isolated experiences, making direct communication impossible without AI acting as intermediary. This would erode the common ground required for meaningful exchange between individuals. At the same time, such private experiences support the growth of personal identity and selfhood. The discussion weighs the risks to social connection against the value of individuality and outlines possible research directions in AI-mediated communication.

Core claim

The paper claims that a scenario of entirely private, AI-generated experiences could lead each person to inhabit a separate experiential world, resulting in individualized languages that remain intelligible to others solely because AI translators stand between them. This intermediation would challenge the integrity of common ground and therefore of communication, while private experience continues to serve as an essential engine of identity and selfhood.

What carries the argument

The Filter Babel thought experiment, which models fully private AI-generated content and encounters that isolate personal experiences and require AI mediation for any cross-understanding.

If this is right

  • Every exchange would depend on the accuracy and neutrality of AI translators, creating new vulnerabilities to error or manipulation.
  • Collective activities that rely on agreed facts or values would face greater obstacles as shared references shrink.
  • Personal identity could develop more freely through customization, yet social cohesion might decline.
  • Research in AI-mediated communication should prioritize designs that sustain both individuality and some form of common ground.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Societal fragmentation could accelerate if common cultural references continue to vanish across groups.
  • AI systems might be engineered to insert occasional shared elements that preserve minimal common ground without removing personalization.
  • Controlled experiments with small user groups exposed to differing personalized feeds could test how quickly private languages emerge.
  • Public institutions like education or news might need deliberate strategies to counteract the isolation of experiences.

Load-bearing premise

A near-future scenario of fully private AI-generated experiences will produce distinct personal languages that others can access only through AI translators.

What would settle it

Observations from users in highly personalized digital environments showing whether they adopt unique terms, references, or concepts that outsiders cannot understand without algorithmic or AI assistance.

read the original abstract

Filter Babel is a thought experiment about a near future in which everything we read, watch, and even whom we "meet" is privately generated for each of us. If we each recede into a world of purely private experience, we may each develop a Wittgensteinian private language that remains intelligible to others only because an AI translator sits in the middle. This intermediation challenges the integrity of common ground and therefore of communication. On the other hand, private experience is an essential engine of identity and selfhood: as Lanier warns, one must be somebody before one can share oneself. This paper opens a discussion of the challenges and opportunities that Filter Babel might present to human communication and identity, and what constructive directions for research in AI-mediated communication might ensue.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper presents the 'Filter Babel' thought experiment, describing a near-future scenario in which AI systems fully personalize all read, watched, and social content for each individual. This personalization is posited to foster Wittgensteinian private languages intelligible to others only through AI mediation, thereby threatening the common ground required for communication. At the same time, the paper emphasizes that private experience remains essential for identity and selfhood, drawing on Lanier to argue that one must be somebody before sharing oneself. The work frames resulting challenges to authenticity and common ground while outlining opportunities and constructive research directions in AI-mediated communication.

Significance. If the envisioned trajectory of extreme personalization materializes, the paper's conceptual framing could help orient HCI and AI research toward preserving shared understanding without eliminating beneficial personalization. It contributes by articulating a clear tension between eroded common ground and the identity-supporting role of private experience, providing a starting point for design-oriented and empirical investigations into synthetic media effects.

minor comments (3)
  1. The abstract would benefit from briefly enumerating the specific research directions proposed later in the manuscript to give readers an immediate sense of constructive next steps.
  2. A short related-work subsection or paragraph situating the Filter Babel scenario against existing empirical studies of filter bubbles, echo chambers, and current recommendation-system effects would help anchor the extrapolation.
  3. The shift from the communication-challenge discussion to the identity-opportunity discussion could be signposted more explicitly to improve flow between the two poles of the central tension.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive and accurate summary of the Filter Babel thought experiment, its framing of tensions between private experience and common ground, and the constructive research directions suggested. The recommendation for minor revision is noted. However, the report lists no specific major comments or points requiring clarification, expansion, or correction.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper presents a conceptual thought experiment on AI-mediated private experiences and common ground, drawing on external references like Wittgenstein and Lanier without any equations, derivations, fitted parameters, predictions, or self-citations that reduce claims to inputs by construction. The central discussion is framed as an opening for research rather than a formal chain that collapses internally, making it self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central scenario depends on untested assumptions about how personalization affects language and identity, with no free parameters or invented entities but reliance on domain assumptions from philosophy.

axioms (2)
  • domain assumption Private experiences are essential for developing individual identity and selfhood (referencing Lanier).
    Invoked directly in the abstract as a counterpoint to the risks of private languages.
  • ad hoc to paper Fully AI-personalized content and interactions will cause individuals to develop private languages intelligible only via AI mediation.
    This is the core premise of the Filter Babel thought experiment.

pith-pipeline@v0.9.0 · 5421 in / 1281 out tokens · 25942 ms · 2026-05-10T08:09:29.675514+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

47 extracted references · 16 canonical work pages

  1. [1]

    Dhruv Agarwal, Mor Naaman, and Aditya Vashistha. 2025. AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuances. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI ’25). Association for Computing Machinery, New York, NY, USA, Article 1117, 21 pages. doi:10.1145/3706598.3713564

    work page doi:10.1145/3706598.3713564 2025

  2. [2]

    Baraniuk

    Sina Alemohammad, Josue Casco-Rodriguez, Lorenzo Luzi, Ahmed Imtiaz Hu- mayun, Hossein Babaei, Daniel LeJeune, Ali Siahkoohi, and Richard G. Baraniuk

  3. [3]

    Self-consuming generative models go mad

    Self-Consuming Generative Models Go MAD. arXiv:2307.01850 [cs.LG] https://arxiv.org/abs/2307.01850

    work page arXiv

  4. [4]

    Bergstrom and C

    Carl T. Bergstrom and C. Brandon Ogbunu. 2024. AI Will Turn Our Lives into The Truman Show.Scientific American(Dec. 2024). https://www.scientificamerican. com/article/ai-will-turn-our-lives-into-the-truman-show/ Opinion

    2024

  5. [5]

    Matyas Bohacek and Hany Farid. 2023. Nepotistically Trained Generative-AI Models Collapse. arXiv:2311.12202 [cs.AI] https://arxiv.org/abs/2311.12202

    work page arXiv 2023

  6. [6]

    Melvyn Bragg. 2020. Authenticity. https://www.bbc.co.uk/programmes/ m00035z4. In Our Time. BBC Radio 4 broadcast 19 November 2020; accessed 2026-02-11

    2020

  7. [7]

    Charlie Brooker. 2023. Joan Is Awful. Television episode, in Black Mirror, season 6, episode 1. https://en.wikipedia.org/wiki/Joan_Is_Awful Directed by Ally Pankiw; premiered on Netflix, 15 June 2023

    2023

  8. [8]

    Stewart Candlish and George Wrisley. 2019. Private Language. InThe Stanford Encyclopedia of Philosophy(Fall 2019 ed.), Edward N. Zalta (Ed.). Metaphysics Research Lab, Stanford University

    2019

  9. [9]

    2010.The Shallows: What the Internet Is Doing to Our Brains(1st ed.)

    Nicholas Carr. 2010.The Shallows: What the Internet Is Doing to Our Brains(1st ed.). W. W. Norton & Company, New York

    2010

  10. [10]

    Matthew Chalmers and Ian MacColl. 2003. Seamful and seamless design in ubiquitous computing. InWorkshop at the crossroads: The interaction of HCI and systems issues in UbiComp, Vol. 8

    2003

  11. [11]

    Matteo Cinelli, Gianmarco De Francisci Morales, Alessandro Galeazzi, Walter Quattrociocchi, and Michele Starnini. 2021. The echo chamber effect on social media.Proceedings of the National Academy of Sciences118, 9 (2021), e2023301118. doi:10.1073/pnas.2023301118

    work page doi:10.1073/pnas.2023301118 2021

  12. [12]

    H., & Brennan, S

    Herbert H. Clark and Susan E. Brennan. 1991. Grounding in communication. In Perspectives on socially shared cognition, Lauren B. Resnick, John M. Levine, and Stephanie D. Teasley (Eds.). American Psychological Association, Washington, DC, Chapter 7, 127–149. doi:10.1037/10096-006

    work page doi:10.1037/10096-006 1991

  13. [13]

    2011.Ready Player One(1st ed.)

    Ernest Cline. 2011.Ready Player One(1st ed.). Crown Publishers, New York

    2011

  14. [14]

    Anna L Cox, Sandy JJ Gould, Marta E Cecchinato, Ioanna Iacovides, and Ian Ren- free. 2016. Design frictions for mindful interactions: The case for microboundaries. InProceedings of the 2016 CHI conference extended abstracts on human factors in computing systems. 1389–1397

    2016

  15. [15]

    Gilles Deleuze. 1992. Postscript on the Societies of Control.October59 (1992), 3–7. http://www.jstor.org/stable/778828

    1992

  16. [16]

    2025.Showrunner

    Fable Simulation. 2025.Showrunner. https://www.showrunner.xyz/ Accessed: 2026-02-09

    2025

  17. [17]

    William W Gaver, Jacob Beaver, and Steve Benford. 2003. Ambiguity as a resource for design. InProceedings of the SIGCHI conference on Human factors in computing systems. 233–240

    2003

  18. [18]

    Kenneth J. Gergen. 1991.The Saturated Self: Dilemmas of Identity in Contemporary Life. Basic Books, London, England

    1991

  19. [19]

    Is model collapse inevitable? Breaking the curse of recursion by accumulating real and synthetic data,

    Matthias Gerstgrasser, Rylan Schaeffer, Apratim Dey, Rafael Rafailov, Henry Sleight, John Hughes, Tomasz Korbak, Rajashree Agrawal, Dhruv Pai, Andrey Gromov, Daniel A. Roberts, Diyi Yang, David L. Donoho, and Sanmi Koyejo. 2024. Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data. arXiv:2404.01413 [cs.LG] ...

    work page arXiv 2024

  20. [20]

    Culture now has two audiences: people and machines

    Giuseppe Granieri. 2014. Algorithmic culture. “Culture now has two audiences: people and machines”: A conversation with Ted Striphas. https://medium.com/futurists-views/algorithmic-culture-culture-now- has-two-audiences-people-and-machines-2bdaa404f643. Published in Futurists’ Views (Medium). Accessed 2026-02-11

    2014

  21. [21]

    Jeffrey T Hancock, Mor Naaman, and Karen Levy. 2020. AI-Mediated Communication: Definition, Research Agenda, and Ethical Considerations. Journal of Computer-Mediated Communication25, 1 (01 2020), 89–100. arXiv:https://academic.oup.com/jcmc/article-pdf/25/1/89/32961176/zmz022.pdf doi:10.1093/jcmc/zmz022

    work page doi:10.1093/jcmc/zmz022 2020

  22. [22]

    Jess Hohenstein and Malte Jung. 2020. AI as a moral crumple zone: The effects of AI-mediated communication on attribution and trust.Computers in Human Behavior106 (2020), 106190. doi:10.1016/j.chb.2019.106190

    work page doi:10.1016/j.chb.2019.106190 2020

  23. [23]

    Kizilcec, Dominic DiFranzo, Zhila Aghajari, Hannah Mieczkowski, Karen Levy, Mor Naaman, Jeffrey T

    Jess Hohenstein, René F. Kizilcec, Dominic DiFranzo, Zhila Aghajari, Hannah Mieczkowski, Karen Levy, Mor Naaman, Jeffrey T. Hancock, and Malte F. Jung

  24. [24]

    Hollenbeck, Charles R

    Artificial intelligence in communication impacts language and social relationships.Scientific Reports13 (2023), 5487. doi:10.1038/s41598-023-30938-9

    work page doi:10.1038/s41598-023-30938-9 2023

  25. [25]

    Maurice Jakesch, Advait Bhat, Daniel Buschek, Lior Zalmanson, and Mor Naaman

  26. [26]

    Co-Writing with Opinionated Language Models Affects Users’ Views. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems Restoring Human Authenticity in AI-Mediated Communication, April 15, 2026, Barcelona, Spain Advait Sarkar (Hamburg, Germany)(CHI ’23). Association for Computing Machinery, New York, NY, USA, Article 111, 15 page...

    work page doi:10.1145/3544548.3581196 2023

  27. [27]

    1984.The Unbearable Lightness of Being

    Milan Kundera. 1984.The Unbearable Lightness of Being. Harper & Row, New York. Originally published asNesnesitelná lehkost bytí(1984)

    1984

  28. [28]

    2010.You Are Not a Gadget: A Manifesto

    Jaron Lanier. 2010.You Are Not a Gadget: A Manifesto. Allen Lane, London. UK hardback edition (Allen Lane imprint)

    2010

  29. [29]

    Xiaoyan Li, Naomi Yamashita, Wen Duan, Yoshinari Shirai, and Susan R. Fussell

  30. [30]

    ACM Hum.-Comput

    Improving Non-Native Speakers’ Participation with an Automatic Agent in Multilingual Groups.Proc. ACM Hum.-Comput. Interact.7, GROUP, Article 12 (Dec. 2022), 28 pages. doi:10.1145/3567562

    work page doi:10.1145/3567562 2022

  31. [31]

    Hancock, Mor Naaman, Malte Jung, and Jess Hohenstein

    Hannah Mieczkowski, Jeffrey T. Hancock, Mor Naaman, Malte Jung, and Jess Hohenstein. 2021. AI-Mediated Communication: Language Use and Interpersonal Effects in a Referential Communication Task.Proc. ACM Hum.-Comput. Interact. 5, CSCW1, Article 17 (April 2021), 14 pages. doi:10.1145/3449091

    work page doi:10.1145/3449091 2021

  32. [32]

    2011.Cognitive capitalism

    Yann Moulier-Boutang. 2011.Cognitive capitalism. Polity

    2011

  33. [33]

    2019.Digital Minimalism: Choosing a Focused Life in a Noisy World

    Cal Newport. 2019.Digital Minimalism: Choosing a Focused Life in a Noisy World. Portfolio/Penguin, New York

    2019

  34. [34]

    2011.The Filter Bubble: What the Internet Is Hiding from You

    Eli Pariser. 2011.The Filter Bubble: What the Internet Is Hiding from You. Penguin Press, New York

    2011

  35. [35]

    Advait Sarkar. 2023. Exploring Perspectives on the Impact of Artificial Intelli- gence on the Creativity of Knowledge Work: Beyond Mechanised Plagiarism and Stochastic Parrots. InProceedings of the 2nd Annual Meeting of the Symposium on Human-Computer Interaction for Work(Oldenburg, Germany)(CHIWORK ’23). Association for Computing Machinery, New York, NY,...

    work page doi:10.1145/3596671.3597650 2023

  36. [36]

    Ilia Shumailov, Zakhar Shumaylov, Yiren Zhao, Nicolas Papernot, Ross Anderson, and Yarin Gal. 2024. AI models collapse when trained on recursively generated data.Nature631, 8022 (2024), 755–759

    2024

  37. [37]

    Paul Skallas. 2021. Refinement Culture. https://medium.com/@lindynewsletter/ refinement-culture-51d96726c642. Published in The Lindy Newsletter on Medium on December 15, 2021; accessed 2026-02-11

    2021

  38. [38]

    Paul Skallas. 2022. Culture is Stuck. https://lindynewsletter.beehiiv.com/p/ culture-stuck. Published in The Lindy Newsletter on December 31, 2022; accessed 2026-02-11

    2022

  39. [39]

    Paul Skallas. 2023. Refinement Culture. https://lindynewsletter.beehiiv.com/p/ refinement-culture. Published in The Lindy Newsletter on January 01, 2023; accessed 2026-02-11

    2023

  40. [40]

    1992.Snow Crash(1st ed.)

    Neal Stephenson. 1992.Snow Crash(1st ed.). Bantam Books, New York

    1992

  41. [41]

    Ted Striphas. 2015. Algorithmic culture.European Journal of Cultural Studies18, 4-5 (2015), 395–412. arXiv:https://doi.org/10.1177/1367549415577392 doi:10.1177/ 1367549415577392

    work page doi:10.1177/1367549415577392 2015

  42. [42]

    2011.Alone Together: Why We Expect More from Technology and Less from Each Other

    Sherry Turkle. 2011.Alone Together: Why We Expect More from Technology and Less from Each Other. Basic Books, New York. Hardcover edition

    2011

  43. [43]

    2012.Brideshead Revisited: The Sacred and Profane Memories of Captain Charles Ryder(revised text (1959) ed.)

    Evelyn Waugh. 2012.Brideshead Revisited: The Sacred and Profane Memories of Captain Charles Ryder(revised text (1959) ed.). Little, Brown and Company. eBook edition distributed by Hachette Digital, Inc

    2012

  44. [44]

    D. W. Winnicott. 1953. Transitional Objects and Transitional Phenomena: A Study of the First Not-Me Possession.International Journal of Psycho-Analysis34, 2 (1953), 89–97. Originally presented to the British Psycho-Analytical Society, 30 May 1951; PubMed PMID: 13061115

    1953

  45. [45]

    1953.Philosophical Investigations(1st ed.)

    Ludwig Wittgenstein. 1953.Philosophical Investigations(1st ed.). Basil Blackwell, Oxford. Bilingual German–English edition (Philosophische Untersuchungen); pages numbered in duplicate; translated by G. E. M. Anscombe

    1953

  46. [46]

    Mika Yasuoka and Pernille Bjorn. 2011. Machine Translation Effect on Communi- cation: What Makes It Difficult to Communicate through Machine Translation?. In2011 Second International Conference on Culture and Computing. 110–115. doi:10.1109/Culture-Computing.2011.28

    work page doi:10.1109/culture-computing.2011.28 2011

  47. [47]

    Yongle Zhang, Dennis Asamoah Owusu, Marine Carpuat, and Ge Gao. 2022. Facilitating Global Team Meetings Between Language-Based Subgroups: When and How Can Machine Translation Help?Proc. ACM Hum.-Comput. Interact.6, CSCW1, Article 90 (April 2022), 26 pages. doi:10.1145/3512937

    work page doi:10.1145/3512937 2022