General positioning

Our research is informed by critical theory and maintains an explicit commitment to sex positivity, including support for pornography and sex work.

We, Anthony Masure and Saul Pandelakis, supervisors of the Fucking Tech! project, unequivocally condemn and prohibit any content involving minors, non-consensual acts, rape or bestiality.

Ethics relating to sexualities

This research-creation project adopts a critical stance towards the normative representations that could potentially emerge in the design of sexbots. Our approach acknowledges that the figure of the robot (and now conversational agents and AI), often conceived as subordinate, tends to be feminized and, consequently, susceptible to perpetuating sexist, heteronormative, racist, and ableist stereotypes.

We set up a framework to ensure the experiments would be conducted in a manner that respects our common ethics and respective boundaries. 

Because such an affirmation may appear performative, in the negative sense of the term (such as, for instance, the claim that a space is a “safe space”), we felt the need to prepare for the 3-year project by asking all the participants in the project about their personal concerns regarding a project that involved nudity, triggering content and NSFW materials. 

An online survey (submitted to the Inclusivity Delegate of HEAD – Genève) was shared with the team members, asking them to rate the sensitiveness of first sexual experiences, sexual assault, sexual harassment, incest, sex work, sextoys, BDSM practices, nudity, group sex, male domination / sexism, white domination / racism, power play, conversion therapy and queer or LGBTphobia / homophobia, transphobia. The survey included a criteria-free zone to add any triggering topics. We then offered common ground rules to organize the work, such as explicit trigger warnings, frequent breaks or the ability to pass on a given activity. 

Using this initial survey, Anthony and Saul then formalized a “Good Practices Guide” that they shared with the participants involved in the project. This document was presented as an open-ended, in-progress tool that could and can still be modified according to the project’s and participants’ needs. While the guidelines affirmed in the guide were not always respected during the project, its presence helped maintain manageable working conditions and brought in an awareness of potential breaches in the projected working contract.

Here are the terms of our guidelines, summerized:

• Flagging sensitive topics.
• Specific framing for the workshops: 
• • participants should be able to pass on an activity at any time;
  • post-activities debriefs will include feedback on the experience and the potential issues encountered.
  • We (Anthony and Saul) can offer one-on-one debrief at any point, IRL or on Zoom if need be on a specific topic.
  • We will do our best to set a time limit to a given activity.
• Content of the StoryTellurX app: dummy content will be provided to replace explicit images, specifically for those testing the app in terms of development, where the content of the images matters less and being repeatedly exposed to it could have nefarious effects.
• Conduct when faced with unintentionally illegal content/violation of our boundaries: the image or text is discussed and analyzed. Extra breaks are offered if needed. If the problematic content is an image, it is doctored to erase all explicit areas. Such pictures are not stored on our personal computers.
• Status of the ‘Good Practices’ document: this document is not a fixed set of rules but a work in progress. It will be edited based on your feedback. A version of this document will exist for our archives, and edited versions will accompany us as the project unfolds to fit your and our specific needs.

Ethics relating to AI & sustainability 

We can clarify our position by referring to the “10 Moral Questions” (2021) by QCollective (thanks to Marc Friederich), a group of citizens and technology users. These questions are divided into 5 “values,” which we outline below with responses related to the Fucking Tech! project and the StoryTellurX sub‑project.

1 – Care & Compassion

Q1 – How is my technology beneficial to the world at large?

In a world saturated with AI Girlfriends (AI GF) of poor conversational quality that perpetuate sexist and racist stereotypes, offering an alternate chatbot featuring sexual content is of the utmost importance. We have observed that the existing services (AI GF or sex dolls) reproduce a cis-heterosexual white and able gaze that actively excludes many folks or subjects them to a limited, disappointing or offensive experience.

Q2 – To what extent is my awareness of our interconnectedness built into this technology so as not to cause anyone or anything suffering?

Wholesale rejection of sexual chatbots through censorship, will neither halt the proliferation nor address their flaws. Moreover, such approaches reflect an anti-sex stance we firmly oppose. There is no reason to consider that chatbots could not be sexual or enable users to have sexual experiences, except if one agrees to separate sexualities from the continuum of life. Using services (such as LLMs) that rely on extractivism, cheap labor and copyrights’ violations is indeed problematic. However, turning away from chatbots does nothing but allow the existing market to carry on its production unchallenged, with its morally questionable foundations and equally dubious outputs. 

2 – Discovery

Q3 – How does my creation nurture the freedom of discovery?

StoryTellurX features a gallery of diverse characters and entities, including non-anthropomorphic forms, moving away from the typical market focus on stereotypical feminine bodies. This commitment to exploring diverse sexual approaches and identities is central to the project.

Q4 – Am I (or is someone) able to shut it off?

StoryTellurX is designed to automatically end conversations after a set number of exchanges, a feature absent from existing AI GF platforms in our comparative analysis. This design choice establishes experiential boundaries while reducing energy consumption, and reflects our theoretical position that desire is constituted through resistance and experiences of absence.

3 – Holistic Thinking

Q5 – What is the worst thing that can happen if my technology is operated with an evil purpose?

One could imagine that the project’s open-source code might serve to reinforce existing commercial offerings and their accompanying political assumptions (this risk exists for any open or free technology). This accessibility could equally enable activist communities to extend and adapt the methodology toward alternative objectives.

Q6 – What will our grandchildren think about the impact of this creation on such things as planetary boundaries, the universe, and the dignity of the earth’s inhabitants?

This is a question we wish to challenge. Drawing from Lee Edelman (No Future, 2004), we are wary of the use of children as an overarching, homogenous category often mobilized to fuel a moral compass. Why do we need to refer to non-existing children to justify our production? If we wish to address an apparatus’ sustainability, we can do so by looking at the current outputs of such production, as Kate Crawford, among others, has done in Atlas of AI (2021).  The constant turn to “the future”, while coming from an ethically sound posture, forecloses the way we think about our actions. It is of course important to make predictions and to analyze if your actions might impact negatively the climate and the social realm. But this imperative must account for both the present and the future, and the very category “future” needs to be unhinged from linear and homogenous understandings of time.

4 – Transparency & Integrity

Q7 – Do I truly understand how my technology works and am I willing to invest in making that transparent in order to educate all stakeholders?

StoryTellurX’s source code is open source and extensively commented. For the data model, we chose compatibility with both open and proprietary models to enable comparative analysis for research purposes. The project’s overall design methodology will be detailed in a dedicated publication. Importantly, we operate without investors as a non-profit initiative funded by Swiss public services.

Q8 – How does my technology creation encourage people to interact with integrity? 

The plasticity and diversity of StoryTellurX’s entities encourage users to question their relationship with multiple forms of sexuality. Consent and respect for personal boundaries and characteristics (such as pronouns) are central to the experience. We actively seek user feedback and will conduct extensive testing before the chatbot’s public release.

5 – Balance

Q9 – Are ethics and sustainability an integral part of my creation?

Yes, for the reasons outlined above. The source code is auditable and hosted on GitHub. Using multiple data models ensures project continuity should any one API become unavailable. If needed, we can also self-host the data server.

Q10 – Have I designed my creation to avoid further inequity and fragmentation of needs or desires?

StoryTellurX doesn’t claim to “solve” existing inequalities, but it can contribute at its own scale by challenging dominant chatbot models and demonstrating that alternative approaches are possible—especially for marginalized communities.