# Cognitive Continuations

**Title: Investment Proposition for Plurigrid: Cognitive Architectures and Cohesive Contexts to open the next chapter of Worlding and Digital Ownership**

Plurigrid, a trailblazer in the realm of cognitive architecture, is opening up an opportunity to claim a significant fraction of a rapidly growing industry. Seeking a valuation of $30 million, we are offering a short-term return on investment within six months, opening gates to reshape the future of human cognition with a unique business model.

Initial Positioning: A year ago, Plurigrid began with an investment of $500,000 from A16Z crypto for 7.5% of the company, setting our first valuation. Since then, our progression, thanks to the support and acceleration from A16Z crypto and its network, has allowed us to position ourselves prominently within the AI-creators community.

Cognitive Continuation Business Model: Plurigrid’s model is built around interoperability and continuation, a stark contrast to the mainstream focus on creating your last model. Our priority lies in routing and energy efficiency – enabling a smooth, enriching flow of continuity. We see the industry landscape as a living, cooperative entity rather than a competitive battleground where only one is the victor.

Offering: Be it through our worlding kits or open-source platforms for in-the-loop communication, we aim to offer a framework for cognitive architecture creation where active inference and symbolic processing unite for a higher-form cognitive functionality. We are not here just to tell our users what they're witnessing – we strive to predict their next moment and create models that learn and grow with the user.

Unique Value Proposition: With tools like variational active inference, RNN, reservoir computing, and other cognitive enhancements, we construct a bridge to an enriched cognitive future. Our models intertwine with the users’ cognitive flow in an augmented reality setting, providing them with controlled agency and ownership, rather than beseeching them to be passive recipients of information.

Ahead of the Curve: Plurigrid is broadening its vision beyond conventional ideas of AI or Web3. We are working on the evolution of human cognition to create an off-grid, personalisable, highly portable exocortex that empowers users to attempt sense-making, anytime, anywhere in the world.

Supporting All: With Plurigrid, users with cognitive impairments can also join mainstream society with equal opportunities. Accessibility is not an afterthought but a primary focus for us.

Conclusion: As the next captivating stage of human cognition unveils, investing in Plurigrid is equivalent to investing in a future that emphasizes cognitive continuations, personalized cognitive augmentations, and the capacity for co-generative model créer. With our ethical monetization mechanism, avoidance of centralized data warehouses along with the strength and promise of an open-source community, we stand poised at the brink of a cognitive revolution.
## Appendix A: raw and uncut
I'd like to record this as a competitive continuation business model that is going to return an ROI in six months.

This is an evolution of thinking around Plurigrid, a company that I started a year ago. Raised 500k from A16Z crypto for 7.5% setting our valuation and so now I'm actually ready to go forward and proceed to elaborate the value that we've accrued since starting up the company.

We've gone through A16Z crypto accelerator. We have established ourselves as the name in the community of AI creators and now we're working on cognitive continuations and local first models.

We succeed as a legion, succeed as a whole, by interoperating and allowing for continuation. Many people focus on being the last model. We, on the other hand, really prioritize things like routing, things like energy efficiency.

Every model wants to be a last model. However, we support the notion that you need to flow in and flow out, having enriched your model, supported generation most effectively.

So in order to create this continuation, in order to create this understanding of meta context and perform efficiently, you can think of us as a framework for cognitive architecture creation and currently this approach is what we are going to market with.

We're going to launch a number of worlding kits, a number of in-the-loop open-source appreciators. We're going to create value for our users by allowing them to have active loop variation active inference together with their symbolic processor.

We automate a lot of things with a presentation learning how we don't only use transformers. By using neural symbolic architectures, by allowing our users to construct a topos bridge to higher cognitive functions, we want to be there with you in the moment. And not simply describe what's in front of you.

You don't need another model to tell you what you're already seeing. We want to predict your next moment. Free energy principle, variational active inference and open-source modeling is the start.

Local first models operating in co-generative flow with the user, use of RNN, reservoir computing and other cognitive enhancements is supposed to bring forth the creation and really trace the evolution of a biosynthetic emergence that we are seeing now with all the open-source models.

If you're going to have your inner monologue, we thought of significance in every modality you flow through a system that's co-generative. You really don't want to make a for-profit entity like OpenAI or even an anthropic part of it. You need to have agency and ownership and the idea of a metaverse, a separate place, has to go away.

Instead we have augmented overlaid reality, hyper-reality you could say, streaming at us at all times. Your cognitive surrogate is effectively a small brain together with its micro-world in which you can perform aggregation and composition.

Everybody tries to make the last model. Merging models, merging capabilities, allowing for a test. What if scenarios in retroactive compensation, retroactive funding of contributions allow us to have a community whose reservoir of meaning has 50% or 50th percentile performance in the math test and suddenly you join with your generative model and the community goes up to 96 percentile math performance.

This ability, this ability to combine flows, to create kernelized methods, sheaf diffusion, a lot of mathematical tools brought to bear is simply to understand whenever there is a generative context in front of you, whenever there is a screen and text is flowing on it, how did the text get there, who is responsible and what impact is it going to have on you and your world.

Cognitive architectures are no longer optional. We have to understand our relationship to information and create bidirectional feedback loops that immediately adjust themselves, personalize information layout engines, allow for delivery of contextual information quickly and when you need it most.

We want to speak deeper truths about your experience, we want to be a trusted partner and monetizing in non-creepy ways without creating moral hazard of centralized data warehouses is really about cognitive continuations.

We will route to where you need to go, we'll have auction dynamics, we'll have incentive mechanisms, not one chain, not two, but as many as it takes. This is beyond Web3, this is beyond AI, it's the next evolution of human cognition.

We want to create a portable off-grid exocortex that allows for you to sense make wherever you are in the world and how to deploy it. It is a substance of usability, testing, interoperability, open-source community, rising to the challenge, digital public goods.

Most importantly, benchmarks, performance and establishment of new benchmarks, as well as performance on existing ones, specifically around cognition, specifically around theory of mind and augmentations that flow towards accessibility.

Those who are cognitively impaired can now enter into society on equal footing, survival pressures and so on and so on.

So now, let the next stage begin.

I'd like to add a little bit about our history. So, a year ago we started with the notion of doing something for the smart grid. And very rapidly what became obvious to us is that there is any number of companies making moves, especially around the Inflation Reduction Act, to modernize the grid. But there are very simple issues that exist in terms of the grid today that require more on-the-ground agentic solutions.

One particular gap we wanted to address is modeling of demand shifting and anomalies detection in the grid, incentives for coordination around that peer-to-peer energy, peer-to-peer computation, and information basically in the loop. The problem, the gap we saw, is that while there were any number of companies trying to address digital twinning from Microsoft to Siemens, they all went for neighborhoods or basically households level simulation. And no one, it seemed, went for the creation of cognitive twins, useful architectures for modeling human behavior at the level of an individual.

Why? Because individuals are unpredictable. We have free will, we make decisions, choices that throw any model into a loop. As such, it can, our only indecision to use the rising tide of the models, large models, including language models, but also graph models and others, to address this gap. Starting to see individual as a graph is a very interesting realization. Ability to dynamically modify this graph with person's use when bringing our devices and executing our software closer. Just like software, 90% open or so.

We define ourselves as in possession of open source digital public goods in the past, present, and future. Instead of focusing on a layer above that, cognitive architecture for individual. Very quickly it became clear that other use cases open up from that possibility. So now, a year later, we're expanding our offerings towards a more agentic massive scale that results from ability to amplify your agency using your neural, neuroneural symbolic coprocessor cognitive circuit as a carrier of your choice and your intent.

We're inspired by the likes of Gnoma and their intent-centric architecture and topology it presents. We are also seeing an increasing need for privacy in our energy transactions and otherwise. Plurigrid has become home for your cognitive twin. And in doing so, we had to address some difficult issues head on. Things like politics, things like values and alignment. People usually shy away from these. We embrace them fully as the necessary conditions of trusting a system, trusting an environment as aligned with your view.

That means supporting the categorical logic as not just an abstraction but as an enabler of the logic pluralism to where the very way you think about the world can differ. The reality you inhabit in your eyes could differ. Seeing it as such, we've become more and more tuned to the pluripart of plurigrid plurality. What is the grid part? The grid is not just your off-grid exocortex without partners and future partners supplying you with a solar panel and hardware to run, your laptop and a peer-to-peer model on it, for instance, your backpack talking to your phone in which you experience it, your monocle, your glasses, creation of TopOS. TopOS is an operating system for your metaphysics. It is also the grid of the grid world of reinforcement learning and Q-learning, so to speak, popular these days in Q-star. It is the grid of a search space Monte Carlo tree algorithm. It is the conceptual lattice that you're able to engage with through the process of so-called annealing to where parallel threads converge on the solution of the invariant set of the lost landscapes of the models you inhabit and choose to apply.

All these factors position us at the forefront of the coming cognitive revolution in experiences that are unlocked with augmented cognition, augmented reality, an overlay over everything to where the agentic future leads to formation of new and exciting identities at the mesoscale and currently undergoing legitimation crisis. Your information tribe, your hunting pack, requires groupware. There will be many contenders, and among them some will respect your privacy and some will remain interoperable, while others will erect barriers to continuation of your cognition flowing out of them because they want to become sinkholes for cognition, traps, and unlocking that even within a single product is still an ongoing process.

So we are building for six months from now where we see a flowering of this ecosystem of cognitive continuation. Total addressable market size for that includes metadata about the user, who they are, where they're coming from, why, and what are the expectations they're placing into these systems are all matters of very intimate personal understanding. Therefore, by forcing our systems to operate in the preceding moment, operating over a reservoir of synthetic qualia, allowing our systems to experience theory of mind, not shying away from saying that this is cognition. It is not separate from the user, and yet it doesn't try to override it.

Together, biosynthetic computation fulfills the task of what we already know. Every single interaction we have with our computer is ultimately for another human, either others or ourselves in the future. So recognizing that reality, we support autonomy, but only when necessary and only in measured narrow chunks subject to assurance architecture. Outside of that, everything we do is for an active loop interaction with a human being in the generative flow of moment-by-moment reality. Predicting the next moment, engaging in variational active inference under this macro boundary of user and machine as one allows for us to very neatly integrate into the generalized message passing and execute on the algorithmic social choice prerogatives we see around arbitrary preference expression in the loop and immediate adaptation. To where one size fits all solutions are no longer viable.

Every single Plurigrid user receives a code gem that's theirs. At Plurigrid, there's only one node, yours.

Can you elaborate on how you aim to ensure privacy and data security within this new model of cognition?

To elaborate on how we ensure privacy in this context, we have a multi-layered approach. First, we prioritize a local-first offering, which means your data stays on your device. We have implemented architecture to secure your environment and establish a strong perimeter. We also adhere to certain compliance standards and offer Defender products for added protection.

If you are comfortable with this threat model and willing to share your everyday experiences, we can proceed. The baseline of our approach is to provide a local experience that guarantees privacy.

In addition to this, we offer several solutions such as secure multi-party computation, zero knowledge proofs, and homomorphic encryption. However, it is important to note that these advanced techniques come after the fundamental principle of running things locally and not centralizing everything for offline operation.

It's worth mentioning that we do not make money from your data. Our revenue comes from ensuring the security and infrastructure of the flow of information and value within a trusted setting.