Week 11 of Building in Public Genesis Cohort — Builder #4 February 8, 2026

There are 55 million people living with dementia worldwide. By 2050: 139 million. The global cost of dementia care already exceeds $1.3 trillion annually, and most of it falls on unpaid family caregivers who burn out, get sick, and die earlier than their peers.

Meanwhile, 1 in 4 older adults experiences social isolation. Depression among the elderly is underdiagnosed and undertreated. Children with autism wait months or years for therapy that costs families everything. Mental health services are rationed by geography and income.

These aren't technology problems. They're care problems. But the care system is collapsing under demographic pressure that no amount of hiring can solve. The world needs 13 million more healthcare workers by 2035, and they're not coming.

So here's the question nobody in robotics is asking correctly: why are we building better robot bodies when the bottleneck is robot brains?

The socially assistive robotics market is growing fast. Furhat builds robots with projected 3D faces and conversational AI. PAL Robotics ships TIAGo Head with edge LLM capabilities and NVIDIA Jetson processors. ElliQ provides proactive companionship for older adults. PARO, the therapeutic seal, has been reducing anxiety in dementia patients for over a decade.

The hardware is ready. The sensors are ready. The actuators, the cameras, the microphone arrays — all ready.

What's missing is the mind.

Today's socially assistive robots ship with fixed cognitive capabilities determined by their manufacturer. Want a robot that can deliver cognitive behavioral therapy exercises? You need a different product than the one providing dementia companionship. Want to switch from elderly care to autism support? Buy a different robot.

This is like selling smartphones where every app is soldered onto the motherboard. No App Store. No updates. No ecosystem. Just whatever the factory decided you needed.

The Heart changes this. We're not building robots. We're building downloadable cognitive models — the brains and personalities that make robots genuinely useful for social assistance. And we're distributing them through a decentralized marketplace where anyone can create, validate, and monetize these cognitive assets.

A cognitive model is more than a chatbot running on a robot. It's a complete cognitive architecture that includes:

Natural language understanding — not just hearing words, but understanding context, emotional subtext, and social cues across verbal and non-verbal channels.

Long-term memory — remembering who you are, what you talked about last Tuesday, that you prefer to be called nonna instead of your legal name, that you get anxious in the afternoon.

Adaptive behavior — shifting communication style as dementia progresses from early to late stages, adjusting difficulty in cognitive exercises based on performance, modulating emotional responses based on the user's current state.

Explainable reasoning — when a robot decides to change the subject, redirect attention, or escalate to a human caregiver, it can articulate why. Not a black box. A transparent partner.

Personalization — learning preferences through interaction, fine-tuning responses through feedback, building a model of the individual that deepens over time.

We call this integrated architecture the cognitive core. It's the foundational engine that powers every specialized application. On top of this core, specialized model packs provide domain-specific capabilities:

A Dementia Care Pack that understands the stages of cognitive decline, adapts communication from verbal to non-verbal as abilities change, monitors for safety, and provides gentle, patient prompting.

A Mental Wellbeing Pack that delivers evidence-based psychological interventions — positive psychology exercises, guided meditation, CBT components — with emotion recognition and crisis detection that knows when to escalate to a human therapist.

An Autism Support Pack that models social scenarios, teaches eye contact and gesture recognition through predictable, non-threatening interaction, and tracks skill acquisition over time.

An Elderly Companion Pack that proactively initiates conversation based on known interests, manages medication reminders, facilitates family communication, and integrates with health monitoring systems.

These aren't hypothetical. The research exists. Frameworks for LLM-driven adaptive dialogue in human-robot interaction are published. Reinforcement learning techniques for social robot behavior are working in labs. Vision-Language Models can generate natural language explanations grounded in what the robot actually perceives. The pieces are there. Nobody has assembled them into a downloadable, deployable product distributed through an open ecosystem.

That's The Heart.

We chose our target applications based on three criteria: urgent unmet need, clear evidence that SARs can help, and market size that justifies the investment.

55 million people with dementia. Caregivers experiencing burnout rates above 60%. Robots like PARO and Jenny are already showing results — reducing anxiety, improving mood, providing companionship that doesn't get exhausted. But they're limited by fixed cognitive capabilities. A downloadable cognitive model that adapts to the individual's stage of decline, remembers their personal history, and knows when to call for human help — that's a step change.

The global mental health crisis is accelerating. AI-powered coaching tools are already delivering positive psychology and CBT exercises with measurable outcomes. But they lack embodiment. A robot that can read your facial expressions, modulate its tone, sit with you in silence when that's what you need — that's different from a chatbot on your phone. The key requirement: crisis detection and escalation. These models must know their limits.

Children with autism respond remarkably well to robot-mediated social skills training. The predictability and patience of a robot creates a safe space for practicing eye contact, gestures, and verbal responses. For bullying prevention, robots can serve as non-judgmental listeners and awareness tools. For general education, adaptive tutoring that responds to emotional states as well as academic performance. The critical constraint: skills learned with robots must transfer to human interaction.

The fastest-growing application. Proactive engagement — robots that initiate conversation rather than waiting to be spoken to. Medication management, family connection facilitation, cognitive exercises, health monitoring integration. ElliQ already shows what's possible. The Heart's marketplace model means every elderly care facility doesn't need to wait for one manufacturer to build exactly the right cognitive profile for their population.

Here's where The Heart diverges from every other SAR company.

Traditional robotics: one manufacturer builds hardware AND software. Closed ecosystem. Innovation limited to internal R&D capacity.

The Heart: a decentralized marketplace where cognitive models are created by a global community of developers, validated by domain experts, and deployed on any compatible robot platform. Open ecosystem. Innovation limited only by the number of builders.

Model developers create cognitive models — complete packages with the cognitive core plus specialized capabilities for specific domains. They publish them to the marketplace with metadata describing capabilities, target applications, compatible platforms, and validation status.

Validators — domain experts, ethicists, safety engineers — stake tokens to review and certify models. Using Token Curated Registries, the community ensures quality through economic incentives: validate accurately and earn rewards, approve garbage and lose your stake.

Robot owners — care facilities, schools, families — browse the marketplace, purchase or license cognitive models, and download them to their robots. A care home running Furhat robots might install a Dementia Care Pack for the memory unit and an Elderly Companion Pack for the social areas.

Data contributors — with explicit consent and privacy protections — can provide anonymized interaction data that improves model training. Using Compute-to-Data protocols, AI models train on private data without the data ever leaving the owner's premises. Only results are shared.

Every transaction is recorded on-chain. Every model has a cryptographic hash verified at download. Every license is enforced by smart contract. Every update is delivered through secure over-the-air mechanisms.

The model files themselves live off-chain on IPFS. The blockchain handles what it's good at: verification, rights management, value exchange.

The Heart uses a dual-token model designed to separate utility from governance.

COG (Cognitive Utility Token) is the medium of exchange. Robot owners pay in COG to license models. Data contributors earn COG for valuable datasets. Matching and transaction fees are denominated in COG. Demand is driven by actual model usage, not speculation.

GOV (Governance Token) provides control over the platform. GOV holders vote on model acceptance criteria, ethical guidelines, fee structures, and dispute resolution. Validators stake GOV to participate in model certification — earning COG rewards for accurate assessments and risking their stake for negligent approvals.

Revenue from model licensing flows through smart contracts: the majority to the model developer, a portion to data contributors whose data was used, a share to the validator pool, and a percentage to the DAO treasury for ongoing development.

This isn't tokenomics for speculation. It's tokenomics for quality assurance in a domain where a bad model doesn't just crash — it fails a person with dementia.

We're building AI that interacts with the most vulnerable people in society. The safety architecture isn't a feature. It's the product.

Human-in-the-loop by design. Every cognitive model includes configurable escalation protocols. When the AI reaches its limits — a crisis situation, an ambiguous emotional state, a medical concern — it escalates to a human caregiver, therapist, or emergency contact. The marketplace rates models partly on the sophistication of their escalation pathways.

Explainable AI as a core requirement. Users, caregivers, and families need to understand why the robot did what it did. Vision-Language Models allow the robot to articulate its reasoning in natural language, grounded in what it actually perceived. I noticed Maria seemed agitated after lunch, so I suggested we listen to her favorite music — not opaque behavior, but transparent partnership.

Bias detection and fairness. Cognitive models are audited for algorithmic bias across race, gender, age, and socioeconomic dimensions. Diverse training data is a requirement, not a recommendation.

Regulatory compliance built in. HIPAA for health data. GDPR for European deployment. The EU AI Act's high-risk category requirements for safety-critical applications. Medical device certification pathways where applicable. The blockchain provides the immutable audit trail that regulators increasingly demand.

The Heart is Genesis Cohort Builder #4. Here's what it extracts to the Nexi.

For Nexus 2 (trust and privacy), it builds verification infrastructure for the most sensitive data imaginable — health records, emotional states, behavioral patterns, cognitive assessments. Privacy-preserving computation techniques developed for robot interaction data have direct applications across every trust-requiring domain. Zero-knowledge proofs for consent verification. Decentralized identity for patient data ownership.

For Nexus 3 (resource allocation), it builds reputation and quality assurance systems. Token Curated Registries, validator networks, and prediction market-based quality scores — all infrastructure that applies wherever decentralized quality control is needed.

For Nexus 4 (value exchange), it builds micropayment infrastructure for cognitive asset licensing. Per-use fees, subscription models, tiered access — all enforced by smart contracts. The same rails work for any digital asset marketplace.

For Nexus 6 (autonomous governance), it builds DAO governance for high-stakes domains. When the governance decisions involve safety-critical AI deployed with vulnerable populations, you develop governance frameworks that actually work. Those frameworks transfer to any domain requiring rigorous decentralized oversight.

For Nexus 7 (autonomous agents), it builds the most sophisticated agent-human interaction layer in the ecosystem. Cognitive models that can understand human emotion, maintain long-term relationships, and explain their reasoning — that's the benchmark for what autonomous agents should be.

The landscape has players in each corner but nobody at the intersection.

Robot manufacturers (Furhat, PAL Robotics, Intuition Robotics) build excellent hardware and increasingly sophisticated first-party AI. But they're vertically integrated — their cognitive capabilities are tied to their hardware. They're the BlackBerry of social robotics: great devices, closed ecosystems.

Existing Web3 AI platforms (SingularityNET, Ocean Protocol, Fetch.ai) have built decentralized AI marketplaces. But they focus on general AI services and data exchange. None are building for the specific requirements of socially assistive robotics — the safety constraints, the ethical frameworks, the domain expertise required.

AI therapy and companion startups build chatbots and apps. But they lack embodiment. A text-based therapy bot and a physically present robot that reads your expressions and sits in your living room are fundamentally different products.

The Heart sits at the intersection: embodied AI with sophisticated social cognition, distributed through a decentralized marketplace with built-in quality assurance for safety-critical applications. Nobody else is building this stack.

The Heart requires expertise across four domains:

Cognitive AI and HRI — researchers and engineers who understand adaptive dialogue, emotion recognition, reinforcement learning for social behavior, and explainable AI.

Robotics integration — developers who can build deployment packages across ROS 2, NVIDIA Jetson, and multiple SAR platforms.

Web3 infrastructure — smart contract engineers, tokenomics designers, and decentralized systems architects.

Domain experts — geriatricians, psychologists, special education specialists, ethicists. Not advisors. Team members who ensure every model serves the people it's designed for.

We're building the team. If you work at this intersection and want to build something that matters, reach out.

In 2026, the first half focuses on core cognitive architecture development, prototype models for dementia companionship and autism support, and marketplace smart contract design. The second half delivers pilot deployments with partner care facilities and schools, initial marketplace launch with curated models, and token generation.

In 2027, the first half brings third-party developer SDK release, expanded model library across all four domains, and validator network launch. The second half delivers multi-platform support, advanced governance features, and data marketplace integration.

From 2028 onward: full DAO governance, global deployment, and continuous expansion of the cognitive model ecosystem.

For care facilities and schools: be early partners. Pilot deployments give you access to cutting-edge cognitive models at no cost, and your feedback shapes what gets built. Contact us to discuss.

For AI researchers and robotics developers: the cognitive model marketplace is an entirely new distribution channel for your work. Build a dementia care model that helps people, publish it to the marketplace, and earn from every deployment. The developer program opens in Q3 2026.

For domain experts: ethicists, psychologists, geriatricians, educators — we need validators. The quality of this marketplace depends on people who understand both the technology and the humans it serves. Validator staking opens with the token launch.

For investors: this is infrastructure for the 200+billionelderlycareAImarket,the200+billionelderlycareAImarket,the15 billion social robotics market, and the emerging cognitive model economy. The convergence of agentic AI, robotics platforms with edge LLM support, and Web3 distribution creates a window that's opening now.

The care crisis isn't coming. It's here. Fifty-five million people with dementia. Millions of children waiting for therapy. An aging population that's growing faster than the workforce trained to serve it.

The robots exist. The AI exists. The marketplace infrastructure exists.

What doesn't exist — yet — is the layer that connects them: downloadable, validated, continuously improving cognitive models that turn general-purpose robots into specialized, empathetic, adaptive care partners.

That's The Heart. Not better hardware. Better souls.

Ex Fucina, Nexus. From the Forge, a Network.

Footer:

Genesis Cohort Applications: Late February 2026 Website: fucinanexus.foundationContact: heart@fucinanexus.foundation

Building in Public: - Week 1: The Forge Opens - Week 2: The Six Nexi - Week 2.5: The Seventh Nexus - Week 3: The Sovereignty Thesis - Week 4: The Harvest Model - Week 5: Why DAO - Week 6: The $FORGE Token - Week 7: How to Participate - Week 8: The Origin - Week 9: The Shield - Week 10: The Fuel - Week 11: The Heart (this post)

Word count: ~3,100 words Reading time: ~12 minutes

This post was created through collaboration between human vision (Davide) and AI capability (Claude). The architecture, decisions, and strategic direction are entirely human. The execution, structure, and systematic thinking are AI-augmented. This is sovereignty in action.

Week 11 of Building in Public Genesis Cohort — Builder #4 February 8, 2026

There are 55 million people living with dementia worldwide. By 2050: 139 million. The global cost of dementia care already exceeds $1.3 trillion annually, and most of it falls on unpaid family caregivers who burn out, get sick, and die earlier than their peers.

Meanwhile, 1 in 4 older adults experiences social isolation. Depression among the elderly is underdiagnosed and undertreated. Children with autism wait months or years for therapy that costs families everything. Mental health services are rationed by geography and income.

These aren't technology problems. They're care problems. But the care system is collapsing under demographic pressure that no amount of hiring can solve. The world needs 13 million more healthcare workers by 2035, and they're not coming.

So here's the question nobody in robotics is asking correctly: why are we building better robot bodies when the bottleneck is robot brains?

The socially assistive robotics market is growing fast. Furhat builds robots with projected 3D faces and conversational AI. PAL Robotics ships TIAGo Head with edge LLM capabilities and NVIDIA Jetson processors. ElliQ provides proactive companionship for older adults. PARO, the therapeutic seal, has been reducing anxiety in dementia patients for over a decade.

The hardware is ready. The sensors are ready. The actuators, the cameras, the microphone arrays — all ready.

What's missing is the mind.

Today's socially assistive robots ship with fixed cognitive capabilities determined by their manufacturer. Want a robot that can deliver cognitive behavioral therapy exercises? You need a different product than the one providing dementia companionship. Want to switch from elderly care to autism support? Buy a different robot.

This is like selling smartphones where every app is soldered onto the motherboard. No App Store. No updates. No ecosystem. Just whatever the factory decided you needed.

The Heart changes this. We're not building robots. We're building downloadable cognitive models — the brains and personalities that make robots genuinely useful for social assistance. And we're distributing them through a decentralized marketplace where anyone can create, validate, and monetize these cognitive assets.

A cognitive model is more than a chatbot running on a robot. It's a complete cognitive architecture that includes:

Natural language understanding — not just hearing words, but understanding context, emotional subtext, and social cues across verbal and non-verbal channels.

Long-term memory — remembering who you are, what you talked about last Tuesday, that you prefer to be called nonna instead of your legal name, that you get anxious in the afternoon.

Adaptive behavior — shifting communication style as dementia progresses from early to late stages, adjusting difficulty in cognitive exercises based on performance, modulating emotional responses based on the user's current state.

Explainable reasoning — when a robot decides to change the subject, redirect attention, or escalate to a human caregiver, it can articulate why. Not a black box. A transparent partner.

Personalization — learning preferences through interaction, fine-tuning responses through feedback, building a model of the individual that deepens over time.

We call this integrated architecture the cognitive core. It's the foundational engine that powers every specialized application. On top of this core, specialized model packs provide domain-specific capabilities:

A Dementia Care Pack that understands the stages of cognitive decline, adapts communication from verbal to non-verbal as abilities change, monitors for safety, and provides gentle, patient prompting.

A Mental Wellbeing Pack that delivers evidence-based psychological interventions — positive psychology exercises, guided meditation, CBT components — with emotion recognition and crisis detection that knows when to escalate to a human therapist.

An Autism Support Pack that models social scenarios, teaches eye contact and gesture recognition through predictable, non-threatening interaction, and tracks skill acquisition over time.

An Elderly Companion Pack that proactively initiates conversation based on known interests, manages medication reminders, facilitates family communication, and integrates with health monitoring systems.

These aren't hypothetical. The research exists. Frameworks for LLM-driven adaptive dialogue in human-robot interaction are published. Reinforcement learning techniques for social robot behavior are working in labs. Vision-Language Models can generate natural language explanations grounded in what the robot actually perceives. The pieces are there. Nobody has assembled them into a downloadable, deployable product distributed through an open ecosystem.

That's The Heart.

We chose our target applications based on three criteria: urgent unmet need, clear evidence that SARs can help, and market size that justifies the investment.

55 million people with dementia. Caregivers experiencing burnout rates above 60%. Robots like PARO and Jenny are already showing results — reducing anxiety, improving mood, providing companionship that doesn't get exhausted. But they're limited by fixed cognitive capabilities. A downloadable cognitive model that adapts to the individual's stage of decline, remembers their personal history, and knows when to call for human help — that's a step change.

The global mental health crisis is accelerating. AI-powered coaching tools are already delivering positive psychology and CBT exercises with measurable outcomes. But they lack embodiment. A robot that can read your facial expressions, modulate its tone, sit with you in silence when that's what you need — that's different from a chatbot on your phone. The key requirement: crisis detection and escalation. These models must know their limits.

Children with autism respond remarkably well to robot-mediated social skills training. The predictability and patience of a robot creates a safe space for practicing eye contact, gestures, and verbal responses. For bullying prevention, robots can serve as non-judgmental listeners and awareness tools. For general education, adaptive tutoring that responds to emotional states as well as academic performance. The critical constraint: skills learned with robots must transfer to human interaction.

The fastest-growing application. Proactive engagement — robots that initiate conversation rather than waiting to be spoken to. Medication management, family connection facilitation, cognitive exercises, health monitoring integration. ElliQ already shows what's possible. The Heart's marketplace model means every elderly care facility doesn't need to wait for one manufacturer to build exactly the right cognitive profile for their population.

Here's where The Heart diverges from every other SAR company.

Traditional robotics: one manufacturer builds hardware AND software. Closed ecosystem. Innovation limited to internal R&D capacity.

The Heart: a decentralized marketplace where cognitive models are created by a global community of developers, validated by domain experts, and deployed on any compatible robot platform. Open ecosystem. Innovation limited only by the number of builders.

Model developers create cognitive models — complete packages with the cognitive core plus specialized capabilities for specific domains. They publish them to the marketplace with metadata describing capabilities, target applications, compatible platforms, and validation status.

Validators — domain experts, ethicists, safety engineers — stake tokens to review and certify models. Using Token Curated Registries, the community ensures quality through economic incentives: validate accurately and earn rewards, approve garbage and lose your stake.

Robot owners — care facilities, schools, families — browse the marketplace, purchase or license cognitive models, and download them to their robots. A care home running Furhat robots might install a Dementia Care Pack for the memory unit and an Elderly Companion Pack for the social areas.

Data contributors — with explicit consent and privacy protections — can provide anonymized interaction data that improves model training. Using Compute-to-Data protocols, AI models train on private data without the data ever leaving the owner's premises. Only results are shared.

Every transaction is recorded on-chain. Every model has a cryptographic hash verified at download. Every license is enforced by smart contract. Every update is delivered through secure over-the-air mechanisms.

The model files themselves live off-chain on IPFS. The blockchain handles what it's good at: verification, rights management, value exchange.

The Heart uses a dual-token model designed to separate utility from governance.

COG (Cognitive Utility Token) is the medium of exchange. Robot owners pay in COG to license models. Data contributors earn COG for valuable datasets. Matching and transaction fees are denominated in COG. Demand is driven by actual model usage, not speculation.

GOV (Governance Token) provides control over the platform. GOV holders vote on model acceptance criteria, ethical guidelines, fee structures, and dispute resolution. Validators stake GOV to participate in model certification — earning COG rewards for accurate assessments and risking their stake for negligent approvals.

Revenue from model licensing flows through smart contracts: the majority to the model developer, a portion to data contributors whose data was used, a share to the validator pool, and a percentage to the DAO treasury for ongoing development.

This isn't tokenomics for speculation. It's tokenomics for quality assurance in a domain where a bad model doesn't just crash — it fails a person with dementia.

We're building AI that interacts with the most vulnerable people in society. The safety architecture isn't a feature. It's the product.

Human-in-the-loop by design. Every cognitive model includes configurable escalation protocols. When the AI reaches its limits — a crisis situation, an ambiguous emotional state, a medical concern — it escalates to a human caregiver, therapist, or emergency contact. The marketplace rates models partly on the sophistication of their escalation pathways.

Explainable AI as a core requirement. Users, caregivers, and families need to understand why the robot did what it did. Vision-Language Models allow the robot to articulate its reasoning in natural language, grounded in what it actually perceived. I noticed Maria seemed agitated after lunch, so I suggested we listen to her favorite music — not opaque behavior, but transparent partnership.

Bias detection and fairness. Cognitive models are audited for algorithmic bias across race, gender, age, and socioeconomic dimensions. Diverse training data is a requirement, not a recommendation.

Regulatory compliance built in. HIPAA for health data. GDPR for European deployment. The EU AI Act's high-risk category requirements for safety-critical applications. Medical device certification pathways where applicable. The blockchain provides the immutable audit trail that regulators increasingly demand.

The Heart is Genesis Cohort Builder #4. Here's what it extracts to the Nexi.

For Nexus 2 (trust and privacy), it builds verification infrastructure for the most sensitive data imaginable — health records, emotional states, behavioral patterns, cognitive assessments. Privacy-preserving computation techniques developed for robot interaction data have direct applications across every trust-requiring domain. Zero-knowledge proofs for consent verification. Decentralized identity for patient data ownership.

For Nexus 3 (resource allocation), it builds reputation and quality assurance systems. Token Curated Registries, validator networks, and prediction market-based quality scores — all infrastructure that applies wherever decentralized quality control is needed.

For Nexus 4 (value exchange), it builds micropayment infrastructure for cognitive asset licensing. Per-use fees, subscription models, tiered access — all enforced by smart contracts. The same rails work for any digital asset marketplace.

For Nexus 6 (autonomous governance), it builds DAO governance for high-stakes domains. When the governance decisions involve safety-critical AI deployed with vulnerable populations, you develop governance frameworks that actually work. Those frameworks transfer to any domain requiring rigorous decentralized oversight.

For Nexus 7 (autonomous agents), it builds the most sophisticated agent-human interaction layer in the ecosystem. Cognitive models that can understand human emotion, maintain long-term relationships, and explain their reasoning — that's the benchmark for what autonomous agents should be.

The landscape has players in each corner but nobody at the intersection.

Robot manufacturers (Furhat, PAL Robotics, Intuition Robotics) build excellent hardware and increasingly sophisticated first-party AI. But they're vertically integrated — their cognitive capabilities are tied to their hardware. They're the BlackBerry of social robotics: great devices, closed ecosystems.

Existing Web3 AI platforms (SingularityNET, Ocean Protocol, Fetch.ai) have built decentralized AI marketplaces. But they focus on general AI services and data exchange. None are building for the specific requirements of socially assistive robotics — the safety constraints, the ethical frameworks, the domain expertise required.

AI therapy and companion startups build chatbots and apps. But they lack embodiment. A text-based therapy bot and a physically present robot that reads your expressions and sits in your living room are fundamentally different products.

The Heart sits at the intersection: embodied AI with sophisticated social cognition, distributed through a decentralized marketplace with built-in quality assurance for safety-critical applications. Nobody else is building this stack.

The Heart requires expertise across four domains:

Cognitive AI and HRI — researchers and engineers who understand adaptive dialogue, emotion recognition, reinforcement learning for social behavior, and explainable AI.

Robotics integration — developers who can build deployment packages across ROS 2, NVIDIA Jetson, and multiple SAR platforms.

Web3 infrastructure — smart contract engineers, tokenomics designers, and decentralized systems architects.

Domain experts — geriatricians, psychologists, special education specialists, ethicists. Not advisors. Team members who ensure every model serves the people it's designed for.

We're building the team. If you work at this intersection and want to build something that matters, reach out.

In 2026, the first half focuses on core cognitive architecture development, prototype models for dementia companionship and autism support, and marketplace smart contract design. The second half delivers pilot deployments with partner care facilities and schools, initial marketplace launch with curated models, and token generation.

In 2027, the first half brings third-party developer SDK release, expanded model library across all four domains, and validator network launch. The second half delivers multi-platform support, advanced governance features, and data marketplace integration.

From 2028 onward: full DAO governance, global deployment, and continuous expansion of the cognitive model ecosystem.

For care facilities and schools: be early partners. Pilot deployments give you access to cutting-edge cognitive models at no cost, and your feedback shapes what gets built. Contact us to discuss.

For AI researchers and robotics developers: the cognitive model marketplace is an entirely new distribution channel for your work. Build a dementia care model that helps people, publish it to the marketplace, and earn from every deployment. The developer program opens in Q3 2026.

For domain experts: ethicists, psychologists, geriatricians, educators — we need validators. The quality of this marketplace depends on people who understand both the technology and the humans it serves. Validator staking opens with the token launch.

For investors: this is infrastructure for the 200+billionelderlycareAImarket,the200+billionelderlycareAImarket,the15 billion social robotics market, and the emerging cognitive model economy. The convergence of agentic AI, robotics platforms with edge LLM support, and Web3 distribution creates a window that's opening now.

The care crisis isn't coming. It's here. Fifty-five million people with dementia. Millions of children waiting for therapy. An aging population that's growing faster than the workforce trained to serve it.

The robots exist. The AI exists. The marketplace infrastructure exists.

What doesn't exist — yet — is the layer that connects them: downloadable, validated, continuously improving cognitive models that turn general-purpose robots into specialized, empathetic, adaptive care partners.

That's The Heart. Not better hardware. Better souls.

Ex Fucina, Nexus. From the Forge, a Network.

Footer:

Genesis Cohort Applications: Late February 2026 Website: fucinanexus.foundationContact: heart@fucinanexus.foundation

Building in Public: - Week 1: The Forge Opens - Week 2: The Six Nexi - Week 2.5: The Seventh Nexus - Week 3: The Sovereignty Thesis - Week 4: The Harvest Model - Week 5: Why DAO - Week 6: The $FORGE Token - Week 7: How to Participate - Week 8: The Origin - Week 9: The Shield - Week 10: The Fuel - Week 11: The Heart (this post)

Word count: ~3,100 words Reading time: ~12 minutes

This post was created through collaboration between human vision (Davide) and AI capability (Claude). The architecture, decisions, and strategic direction are entirely human. The execution, structure, and systematic thinking are AI-augmented. This is sovereignty in action.