{"slug": "ai-memory-is-becoming-portable-continuity-still-isn-t", "title": "AI Memory Is Becoming Portable. Continuity Still Isn't.", "summary": "Google now lets eligible Gemini users import memory-like context and chat history from other AI platforms, including ChatGPT and Claude, marking progress in AI portability. However, an engineer argues that export is not true portability, and portability is not continuity, as archives fail to recreate the working environment, settings, and current state of long-running projects.", "body_md": "*Moving your chats is progress. Moving your AI life is a different problem.*\n\nAI services are finally starting to make switching less painful.\n\nGoogle now lets eligible Gemini users import memory-like context and chat history from other AI platforms, including exports from ChatGPT and Claude.\n\nThat is real progress.\n\nFor a long time, “portability” often meant little more than downloading a ZIP file and keeping it somewhere safe. Now a major AI provider is treating another provider’s history as something that can become useful context again.\n\nBut that raises a more important question:\n\n**What exactly has moved?**\n\nChatGPT can export your history and account data.\n\nClaude can export conversation and user data.\n\nGemini can export chats, generated media, uploads, Gems data, and other activity through Google Takeout.\n\nThose features matter.\n\nBut an archive is not a working AI environment.\n\nOpenAI’s own guidance for moving conversations between personal ChatGPT accounts makes this distinction clear. You can upload an exported `conversations.json`\n\nfile into a new chat and use it as reference material.\n\nBut OpenAI explicitly says this is not a full migration.\n\nIt does not recreate the original chats, sidebar structure, settings, memories, GPTs, files, subscriptions, or workspace access.\n\nYou can carry the record of the past without recreating the environment that produced it.\n\n**Export is not portability. And portability is not continuity.**\n\nI think AI portability is easier to understand if we separate four layers.\n\nYou can get your data out.\n\nJSON, ZIP, HTML, media files, or another documented archive format.\n\nThis protects against total lock-in, but by itself it is closer to backup than migration.\n\nAnother system can accept that exported material.\n\nGemini’s new cross-platform import flow is an important example.\n\nThe imported material remains useful in the destination.\n\nIt can be searched, referenced, or reused as context instead of sitting there as an inert archive.\n\nThis is where the industry is starting to make real progress.\n\nYou can change the model, provider, application, runtime, or machine and still resume from where you were.\n\nNot merely recover what you once said.\n\nThat is a different engineering problem.\n\nImagine someone who has spent a year building software with an AI assistant.\n\nThey successfully move ten thousand old conversations to a new provider.\n\nThat is valuable.\n\nBut does the new system now know:\n\nSome of those answers may exist somewhere inside the transcript.\n\nThat is not the same as having current state.\n\nA transcript tells a new model what happened.\n\nA continuity layer must also preserve what remains true now, what has been superseded, where a claim came from, and what work is still open.\n\n**Chat history is a record of the past. Continuity is the ability to resume from it.**\n\nBrowser state is a simple example.\n\nWhen OpenAI ended Atlas, users were told to save important Atlas data before shutdown. Bookmarks would not transfer automatically, open tabs might not transfer, and browser history might not transfer.\n\nMeanwhile, ChatGPT conversation history was separate.\n\nThat illustrates a broader point:\n\nThe state of an AI-assisted task does not live only in the transcript.\n\nIt may also live in:\n\n**Chat history is not work state.**\n\nPortable memory is important, but memory alone is still too narrow.\n\nA memory like:\n\nThe user prefers concise answers.\n\nis useful.\n\nBut compare that with:\n\nApproach B replaced approach A because of constraint X. The decision was accepted in issue 123, and later implementation assumes B unless explicitly superseded.\n\nThat is not just “memory.”\n\nIt is part of project authority and decision state.\n\nA durable AI environment may need to preserve:\n\nRemembering who the user is and knowing where the user currently is in a long-running body of work are related problems.\n\nThey are not the same problem.\n\nGemini’s importer deserves credit because it reduces switching cost.\n\nBut look at the architecture:\n\n```\nChatGPT / Claude\n       ↓\n     Gemini\n       ↓\n  Provider C\n```\n\nChatGPT and Claude data become portable into Gemini because Google built the importer.\n\nIf the user later moves again, Provider C needs to understand enough of the previous provider’s exports to make that migration useful.\n\nThis is much better than no portability.\n\nBut it is still provider-to-provider portability.\n\nA different architecture is possible:\n\n```\n         User-owned continuity\n memory / decisions / projects / state\n                    ↓\n                  doll\n        ↙           ↓           ↘\n    Local AI     ChatGPT     Gemini / Claude / ...\n```\n\nIn this model, the canonical copy of durable state does not live inside whichever AI service happens to be active today.\n\nThe provider is a capability layer.\n\nIt receives the context it needs, does the work it is good at, and can be replaced later.\n\nThe continuity layer remains under the user’s control.\n\nIf an import format changes, an adapter can change.\n\nIf a model is discontinued, another model can be used.\n\nIf cloud access disappears, the durable state does not disappear with it.\n\nThis distinction is central to [doll](https://doll.badjoke-lab.com/).\n\ndoll is not trying to become the foundation model that survives forever.\n\nIt is trying to make the foundation model replaceable.\n\nThe continuity layer belongs on the user’s side:\n\nThe capability layer can change:\n\nThe principle is simple:\n\n**Own the continuity. Rent the capability.**\n\nNone of this means continuity is solved.\n\nIt creates harder questions:\n\nWhat deserves to become durable memory?\n\nHow should stale information be superseded?\n\nHow should conflicting records be resolved?\n\nWhich state is authoritative and which state is only evidence?\n\nHow should continuity adapt to models with different context windows, tools, safety boundaries, and capabilities?\n\nHow much provider-specific state can be translated before the translation becomes fiction?\n\nThose questions are harder than exporting a ZIP file.\n\nThat is exactly why portability and continuity should not be treated as synonyms.\n\nThe direction of the industry is encouraging.\n\nChatGPT, Claude, and Gemini all provide meaningful ways to export user data.\n\nGemini now goes further by importing memory-like context and full chat histories from other AI platforms.\n\nOpenAI provides a reference-based way to reuse exported conversations in another personal ChatGPT account.\n\nThis is not evidence that portability has failed.\n\nIt is evidence that portability has become important enough for major providers to build it.\n\nThe next question is whether users can own the layer above it.\n\n**Your AI should not merely let you export the past. It should let you continue from it.**\n\nGoogle — Import from other AI platforms to Gemini Apps\n\n[https://support.google.com/gemini/answer/16868299?hl=en](https://support.google.com/gemini/answer/16868299?hl=en)\n\nOpenAI — Exporting your ChatGPT history and data\n\n[https://help.openai.com/en/articles/7260999-how-do-i-export-my-data](https://help.openai.com/en/articles/7260999-how-do-i-export-my-data)\n\nClaude Help Center — Export your Claude data\n\n[https://support.claude.com/en/articles/9450526-export-your-claude-data](https://support.claude.com/en/articles/9450526-export-your-claude-data)\n\nGoogle — Download your Gemini Apps data\n\n[https://support.google.com/gemini/answer/16920332?hl=en](https://support.google.com/gemini/answer/16920332?hl=en)\n\nOpenAI — Transfer exported conversations between ChatGPT accounts\n\n[https://help.openai.com/en/articles/9106926-transferring-conversations-from-1-chatgpt-account-to-another-chatgpt-account](https://help.openai.com/en/articles/9106926-transferring-conversations-from-1-chatgpt-account-to-another-chatgpt-account)\n\nOpenAI — Evolving Atlas into ChatGPT for browser-based agentic work\n\n[https://help.openai.com/en/articles/20001371](https://help.openai.com/en/articles/20001371)\n\nOriginally published on the official doll site:\n\n[https://doll.badjoke-lab.com/notes/ai-memory-portability-continuity/](https://doll.badjoke-lab.com/notes/ai-memory-portability-continuity/)", "url": "https://wpnews.pro/news/ai-memory-is-becoming-portable-continuity-still-isn-t", "canonical_source": "https://dev.to/badjoke-lab/ai-memory-is-becoming-portable-continuity-still-isnt-1bk6", "published_at": "2026-08-12 10:36:27+00:00", "updated_at": "2026-08-12 10:47:10.310773+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-products", "ai-infrastructure"], "entities": ["Google", "Gemini", "ChatGPT", "Claude", "OpenAI", "Google Takeout"], "alternates": {"html": "https://wpnews.pro/news/ai-memory-is-becoming-portable-continuity-still-isn-t", "markdown": "https://wpnews.pro/news/ai-memory-is-becoming-portable-continuity-still-isn-t.md", "text": "https://wpnews.pro/news/ai-memory-is-becoming-portable-continuity-still-isn-t.txt", "jsonld": "https://wpnews.pro/news/ai-memory-is-becoming-portable-continuity-still-isn-t.jsonld"}}