{"slug": "why-im-building-doll-a-personal-ai-continuity-system", "title": "Why I’m Building “doll”: A Personal AI Continuity System", "summary": "A developer is building 'doll', an open-source personal AI continuity system designed to preserve user state independently of models, providers, or applications. The system aims to ensure that personal AI environments remain usable even when underlying services change or disappear, by treating models as replaceable reasoning engines and user state as the durable core. Currently in pre-alpha, doll focuses on defining model-independent foundations before connecting any model runtime.", "body_md": "High-performance AI is available to far more people than it used to be.\n\nBut there is no guarantee that access will continue under the same conditions.\n\nPrices may change. Usage limits may become more restrictive. Models may be discontinued. Accounts, regional restrictions, regulation, provider policies, or service shutdowns may make an AI system unavailable even if it worked the day before.\n\nThis is not an argument about whether any particular company should or should not be trusted.\n\nThe problem is that the continuity of a personal AI environment currently depends on conditions the user does not control.\n\nThat is why I started building **\"doll\"**, an open-source personal AI continuity system.\n\ndoll is not a new foundation model. Nor is it intended to replace model runners or interfaces such as Ollama, llama.cpp, Open WebUI, or LM Studio.\n\nIts purpose is to preserve the parts of a personal AI environment that should remain usable even when models, providers, applications, runtimes, interfaces, or machines change or disappear.\n\nAI systems are often described as though the model were the center of everything.\n\nBut the model itself is not necessarily the part a person most needs to preserve.\n\nThe durable parts may include:\n\nModels can be replaced.\n\nA more capable model may become available. A provider may retire an older one. A user may need to move to a smaller local model because the available hardware has changed.\n\nChanging the model should not automatically mean losing the state the user has accumulated.\n\nThe central idea behind doll is:\n\nModels are replaceable reasoning engines. The user's state is the durable core.\n\nRunning a model locally is an important way to reduce external dependency.\n\nA local model may remain usable when a cloud service is unavailable. It can eliminate ongoing API costs and allow data to remain on hardware controlled by the user.\n\nBut local execution alone does not guarantee continuity.\n\nWhat happens if conversations and memory are stored inside an application-specific database belonging to a single local AI application?\n\nWhat happens if that application is no longer maintained, changes its storage format, stops working on a new operating system, or cannot be moved to another machine?\n\nA local application can become another form of lock-in.\n\nFor this reason, doll does not treat the format of ChatGPT, an OpenAI-compatible API, Ollama, Open WebUI, or any other provider, runtime, or interface as the canonical representation of user state.\n\nData imported from another AI environment must be mapped into a documented and inspectable representation.\n\nThe result of that mapping must also be honest about loss.\n\nAn import or export may be:\n\nInformation that cannot be transferred should not silently disappear.\n\nOne of doll's governing principles is:\n\nLocal-complete, cloud-optional.\n\nThe local system should remain useful without API keys, account registration, or a permanent internet connection.\n\nCloud models may eventually be used as optional performance extensions. In many cases, they will be more capable than the models a user can run locally.\n\nBut a cloud service must not become the source of truth for:\n\nLosing access to the cloud may reduce performance.\n\nIt should not erase the user's state or remove the path to recovery.\n\nAt the time of writing, doll is **pre-alpha**.\n\nNo model runtime is connected yet, and it is not a daily-use AI assistant.\n\nThat is intentional.\n\nMany AI projects begin by connecting a model and building a chat interface. doll is taking a different path because the boundaries a model must not cross need to be defined and enforced before the model is introduced.\n\nThe model-independent foundation currently being built includes:\n\nA model is not automatically trustworthy simply because it runs locally.\n\nLocal applications, plugins, imported conversations, documents, search results, and tool output must not gain authority merely because they have been placed into a model's context.\n\nAdding these controls only after connecting a model would risk turning them into secondary protections around a system whose authority model was never clearly defined.\n\ndoll is therefore building the boundary first.\n\ndoll is not intended to be:\n\nMoving to another model may change response quality, behavior, personality, or reasoning ability.\n\nMoving to less capable hardware may reduce what the system can do.\n\nContinuity does not mean that performance never changes.\n\nIt means preserving what should survive those changes:\n\ndoll is not a finished product.\n\nIt cannot yet be installed and used as a complete personal AI assistant.\n\nThe current focus is on establishing model-independent state management, transfer, backup, restoration, portability, and safety boundaries before local model integration begins.\n\nDevelopment is divided into small specifications, issues, branches, and pull requests.\n\nThe repository records not only what has been implemented, but also the reasoning behind the implementation order, its acceptance requirements, and the expected recovery behavior when something fails.\n\nProject website:\n\nSource code, specifications, architecture decision records, and implementation history:\n\n[https://github.com/badjoke-lab/doll](https://github.com/badjoke-lab/doll)\n\nThe goal at this stage is not to attract a large number of users.\n\nRight now, criticism of the design is more useful than promotion.\n\nI am particularly interested in questions such as:\n\nI would also like to learn about existing standards, projects, research, and documented failures that address the same problems.\n\ndoll is still at an early stage.\n\nThat is precisely why I want to define what must survive, what should remain replaceable, and where authority must stop before presenting it as a finished AI system.\n\n*Disclosure: This article was prepared with AI assistance and was reviewed, edited, and approved by the project maintainer.*", "url": "https://wpnews.pro/news/why-im-building-doll-a-personal-ai-continuity-system", "canonical_source": "https://dev.to/badjoke-lab/why-im-building-doll-a-personal-ai-continuity-system-1a1c", "published_at": "2026-06-19 15:58:36+00:00", "updated_at": "2026-06-19 16:07:25.142953+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-safety", "ai-infrastructure", "developer-tools"], "entities": ["doll", "Ollama", "llama.cpp", "Open WebUI", "LM Studio", "ChatGPT", "OpenAI"], "alternates": {"html": "https://wpnews.pro/news/why-im-building-doll-a-personal-ai-continuity-system", "markdown": "https://wpnews.pro/news/why-im-building-doll-a-personal-ai-continuity-system.md", "text": "https://wpnews.pro/news/why-im-building-doll-a-personal-ai-continuity-system.txt", "jsonld": "https://wpnews.pro/news/why-im-building-doll-a-personal-ai-continuity-system.jsonld"}}