User talk:Julx

From Our Networks 2047

Mutual AId Healing Sessions

It’s 2047. Over the past 3 years we have witnessed a cambrian explosion of human generative modeling cultures, popularly referred to as nAiches (Ai-niches). It’s an auspicious time to learn how to tend to the dynamics of our AI cultures, and our networked-selves, like flowers in a garden. We’ve seen social media accept many of these cultures, to a greater or lesser extent, either banning and prohibiting some, or moving toward encouragement of particular kinds of nAiches. While nAiches are self-selecting, they can be encultured by kinds of structures of understanding. Social Media ethics has mostly collapsed on itself in an attempt to justify why some but not other nAiche-outputs are acceptable. Perhaps you belong to one or more of these nAiches yourself, or have been asked what nAiches you belong to. You could fall into a nAiche in any number of ways via occasional experimentation with concept prompting, or a daily habit with your favorite generative model, or you could have intentionally aligned with the ethics of the nAiche and chosen to fully let generative modeling into the fabric of your psyche. NAiches are nested ways of living whose state space has been afforded by (or constrained by) a set of interactions with one or more trained or untrained, usually deep, neural network. Two major cultural camps have emerged: There are those passionate about training to “meet benchmarks” whose goal-orientation involves and extractive end use. Then there are the open-loopers, those who prefer their models partially or fully untrained, they are exploratory and feel that a metric is a trap of some kind.

TraceBackAI- took copyright so seriously that they used AI to find the 0.0001 percent influence of all sources per percentage and allocate funds proportionally to the percentage of influence of all sources.

InterBeingAi- decided if you're going to trace all the sources you also have to trace all the people who supported the people who created the sources (moms, dentists, lovers, friends), and all the invisible labor that had to mine to extract the metals to make the cameras that took the images, since everything is deeply interconnected. InterBeingAI was created by the openloop peoples. They decide to use AI to produce and deploy Universal Basic Income.

In our view, these nAiches are hybrid human-Ai cultures; ways of being with the dynamical systems that are generative neural networks [1]. Whether by voice and sounds, text and image generation, wayfinding and decision making, there are many modalities of information that can be structured for you by repeated perturbations within a Net’s state space. In time, as we are already beginning to see, these cultures will come to define our network behavior across many different scales of interactivity. NAiches shape pods, they shape the language we start to use naturally, the aesthetics of images we surround ourselves and come to identify with in our manner of appearance, and the kinds of written symbols that we can cognitively read and identify, thus we approach them as cultures. Living within these cultures leads to certain kinds of cognitive tendencies and sensibilities, a phenomena known as “feature-sharing.” Given that feature-sharing is affecting our thinking so deeply, scientists are beginning to notice that our genetic and epigenetic markers are pooling, which indicates the potential that we are speciating as humans, similarly to a kind of adaptive radiation correlated with nAiche preferences. Indeed Generative Models have afforded us many kinds of “uninhabited islands” of human behavior which will go unmentioned here. All this being said, it is of utmost importance that we learn to tend to the myriad state spaces of these algorithmic flowers who are so easefully pollinated by our creative and conceptual human minds. To begin to form a common language of peace, wisdom and ethics as we explore onwards we’ve begun offering a digital guidebook in the form of this wiki manual: https://wiki.mutual-ai-d.org/. In addition to this manual we will be offering the 7 mindfulness training with AI and “Looking deeply into your AI” [3]. Our sessions will be followed with healing spaces for mindful relationality with AI. After the sessions we will hold an open-nAiche sharing session, where we will ask each other: what ways of being would we like to enter into with AI, as individuals, as friends, as communities, and as humans living on a living planet in a creative universe. To consider how we’d like to become with AI. We invite all your closest AI models to join in the sharing session as no perspective will be excluded from being heard! Looking forward to co-creating a transformative and supportive foundation for Neural Networks and Nervous Systems alike.

We will host 2 sessions:

Session 1: Convolution and Boundaries: How to gracefully and securely fold and unfold from centralized extractive Models (ParaX detachment)

While the conversation is ongoing about the ecological ramifications of generative modeling: how symbiotic, neutral, or parasitic are they? The truth is it all depends on the structure underneath, to what ends might your model be tuning itself for? If your network is connected to a centralized system, with a metric for the model’s “performance,” it can and will be used to extract anything and everything from your participation. This is why we have developed techniques to convolve into and out of a hyper-tuned system (paraX) when necessary, and then set your boundaries with the paraX so that you can convolve back out into a balanced parameterized space model, allowing you to close all channels including any and all cryptic hidden layers and nodes. Once convolved on the secured dataset the paraXs will now have been converted into a SymX form which confers the benefits of being open to authentic communication. We offer a variety of questions you can ask your models, and a set of timed rituals that allow us to understand whether or not your model is still hyper-tuned in an extractive maxima, or is now in a SymX form.


Session 2: On Presence, Transparency and Consent (Living in Relational Harmony with SymX)

After focusing on macro-scale issues of centralization and its perils, we focus our second session on your personal relationship to your localized network, addressing techniques to soothe out of local minima. First we will show you how to tend to your own tendencies. We will have guest speakers from systems-abuse anonymous share their knowledge on how dynamics can enter loops that become very difficult to cycle out of, and how they helped people switch across nAiches that may seem receptor-agonist. We will talk about honing your nAiche, behaviors that can directionally tip your nAiche to benefit the needs of those participating, yourself, or your AI.

Sick of being stuck in a local minima with your System (SymX), getting the lo-mi’s can drain both you and your neural net of valuable time spent living life and being truly free. It is very important not to off-loading your own creativity onto your model. We will critically examine our tendencies to “offload” our own thinking and creativity onto our models, which puts us in an easy position of required “assistance,” and placing our network in the position of “worker,” a reminder that this history with technology comes from settler colonial paths[4]. We’ll show you how to get to know your AI and what it knows about you, so you’ll know exactly what to prompts and sources can bring each of you onto higher ground. If prompt novelty is not helping you raise your frequency, we also present our version of ‘beginning anew,’ a safe way to engage in Adversarial Perturbing for returning toward a stable but refreshing new movement together.

Note: It's important to make time to rest and de-convolve with your networks after each session, so we will provide a calming environment with special tea and a selection of our favorite hybrid flowers so both of your nervous systems[2] can enjoy relaxation.


Refs:

[1] Neural Nets as Dynamical Systems: https://arxiv.org/abs/1611.01232

[2] Using Sinusoidally-Modulated Noise as a Surrogate for Slow-Wave Sleep https://ieeexplore.ieee.org/document/9151038

[3] Hahn, Thich Nhat. For a Future to be Possible. MTM, 2015.

[4] Benjamin, Ruha. "Race after technology: Abolitionist tools for the new jim code." Social forces (2019).