498 results found

1. Feature Request: "Amnesia Bridge" – Cross-Chat Context Retrieval

   When chats close due to resource limits, all invested context is lost. Users must re-upload files and start from scratch—wasting time and compute.

   Proposal: Let users selectively pull context from older chats into active conversations: "Lumo, check our chat from last week for what we found about XY." -> the desired feature is similar to "websearch" - just scoping the own (internal) activities.

   Optional: User tagging of important chats for quicker discovery.

   Why it fits: Same privacy model, just extended. Decryption stays client-side, metadata stays user-controlled and encrypted. No server-side indexing.

   Impact: Better session continuity, no redundant uploads, preserves invested work.

   When chats close due to resource limits, all invested context is lost. Users must re-upload files and start from scratch—wasting time and compute.

   Proposal: Let users selectively pull context from older chats into active conversations: "Lumo, check our chat from last week for what we found about XY." -> the desired feature is similar to "websearch" - just scoping the own (internal) activities.

   Optional: User tagging of important chats for quicker discovery.

   Why it fits: Same privacy model, just extended. Decryption stays client-side, metadata stays user-controlled and encrypted. No server-side indexing.

   Impact: Better session continuity, no redundant uploads, preserves invested… more

   2 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   1 comment  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

2. Enhance Lumo with Dedicated First-Level Support Training & Automatic Knowledge Updates

   Perhaps it's a generally silly idea, but I often try Lumo as the first point of contact for support regarding Proton products. While the AI assistant delivers good to excellent results in many areas of knowledge, it surprisingly seems to be weak in this area, and I have already had to contend with some misinformation. Therefore, I propose extending Lumo’s capabilities to serve as a reliable, up-to-date first-line support agent. The core idea is to provide Lumo with a focused training set, encompassing, among other things, Proton-specific FAQs, troubleshooting guides, and policy documents, combined with an automated pipeline that feeds the latest developer releases and support team updates directly into the model. Why it matters: Users get instant answers without waiting for a ticket, and routine queries are handled by Lumo, freeing the live support team to focus on complex or security-critical cases.

   Perhaps it's a generally silly idea, but I often try Lumo as the first point of contact for support regarding Proton products. While the AI assistant delivers good to excellent results in many areas of knowledge, it surprisingly seems to be weak in this area, and I have already had to contend with some misinformation. Therefore, I propose extending Lumo’s capabilities to serve as a reliable, up-to-date first-line support agent. The core idea is to provide Lumo with a focused training set, encompassing, among other things, Proton-specific FAQs, troubleshooting guides, and policy documents, combined with an automated pipeline that feeds… more

   6 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   6 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

3. Allow per-project memory with general chat memory disabled

   Enable per-project memory while keeping general chat memory off. This is ideal for users who use general/ghost chats as a scratchpad and want to have memory active only for project chats within their folders.

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

4. Share Lumo Response

   Most other LLM apps have a share icon at the end of a response. When click, a link is generated that points to the response in the conversation. Proton can implement this feature and use the link permissions that exist in Proton Drive. A user could set a default permission policy in the app.

   I use this feature extensively to combat misinformation shared with me by asking Grok, for example, "Is this true?" along with a URL. I then share a link to my AI response result as a source of my research.

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

5. Chat Branching

   Google's Gemini lets you create a branch off of any of the LLM's responses. This allows you to take the conversation in multiple different directions.

   Gemini just creates a new chat for each branch, but I would appreciate if Lumo made the branch chats a child of the original chat for organizational purposes.

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

6. Implement Bias Mitigation for Chinese Open-Weight Models (Qwen, Kimi K2)

   As noted in Lumo's Privacy Documentation, the system currently routes queries to several open-weight models, including Qwen and Kimi K2. While utilizing these models supports cost-efficiency and open-source diversity, research (e.g., NDSS 2026, "S1761") indicates they contain significant internal censorship and narrative engineering aligned with authoritarian state interests.

   Proton’s core mission is to champion digital rights, encryption, and freedom against state surveillance. However, deploying models with embedded propaganda or censorship mechanisms creates a contradiction:

   • Cognitive Risk: Users may receive subtly altered information or suppressed viewpoints on sensitive geopolitical topics (e.g., HK, Taiwan, human rights).
   • Value Misalignment: This conflicts with Proton's historical stance supporting activists and resisting authoritarian control.

   Proton should implement a bias mitigation layer specifically for these models. This could involve:

   • Post-Training Alignment: Applying techniques similar to those discussed in ACL 2023 ("Mitigating Bias in Large Language Models") to strip state-aligned censorship.
   • Routing Logic: Automatically routing sensitive queries (political, human rights, censorship-related) to models with stronger neutrality guarantees (e.g., Apertus, Llama variants) or applying strict red-teaming filters.
   • Transparency: Clearly labeling when a response originates from a model known to have censorship constraints.

   Why This Matters: For a privacy-first platform serving a global audience, the integrity of the information provided is as critical as the encryption of the channel. Mitigating these biases ensures Lumo remains a tool for liberation rather than an unwitting vector for state narratives.

   As noted in Lumo's Privacy Documentation, the system currently routes queries to several open-weight models, including Qwen and Kimi K2. While utilizing these models supports cost-efficiency and open-source diversity, research (e.g., NDSS 2026, "S1761") indicates they contain significant internal censorship and narrative engineering aligned with authoritarian state interests.

   Proton’s core mission is to champion digital rights, encryption, and freedom against state surveillance. However, deploying models with embedded propaganda or censorship mechanisms creates a contradiction:

   • Cognitive Risk: Users may receive subtly altered information or suppressed viewpoints on sensitive geopolitical topics (e.g., HK, Taiwan, human rights).
   • Value Misalignment: This conflicts with Proton's… more
   5 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

7. Option turn off anthropomorphic thinking and text

   I am currently very ritated

   why? I am wasting time, energy, and most of all TOKENS because of lumo's anthropomorphic beheviour.

   I understand people love anthropomorphic AIs.

   Can we however have an option to turn if off?

   And i really really really hope i do not need to explain what anthropomorphic in Ais means.

   because i my suggestion is NEXT LEVEL

   as in thinking and the text.

   Lumo keeps repeating the same answers, or otherwise, without thinking even when explicitly told to stop.

   when it finally stops to thinking, more then half of the "thinking" are simple anthropomorphic related text.

   Oh User is frustrated because i did this and i should have not. I need to make sure i apologies to user. I need to make sure i acknolage. I need to this. I need to do that. <- absolutly NONE of them are relevant to the task. and albsolutly ALL of them occure each time Lumo finally thinks.

   We are wasting most of out time, energy, and resources, because of this anthropomorphic elements in an AI.

   or what, is it a crime, heath risk, fraud, abuse, something? that an ais is not constanly determin how to "not offend" someone who is explicitly telling it to stop with this commenteries about what it did, what it did wrong, apologies, acknolagements, what it will do, as in reassurances, and ..... Did i make it clear? ALL of this, even if they are NOT in the output, they are if you simple check the thinking. And this happens alll the time. Insted of Lumo chekcing the tast, or instructions, or otherwise, it preocupised with anthropomorphic beheviour.

   For proton: I hope you understand that I am not talking about preferences but about the fact that a paid product is wasting time, mental energy, is getting absurd for not reason, it's producing text that is unsolicited, unproductive, ilogical at time AND it is PREVENTING anything to progress:

   You are right, i should have not, i will from now

   Here is again what i just wrote because I run out of tokens or time to actually proccess the task

   I am currently very ritated

   why? I am wasting time, energy, and most of all TOKENS because of lumo's anthropomorphic beheviour.

   I understand people love anthropomorphic AIs.

   Can we however have an option to turn if off?

   And i really really really hope i do not need to explain what anthropomorphic in Ais means.

   because i my suggestion is NEXT LEVEL

   as in thinking and the text.

   Lumo keeps repeating the same answers, or otherwise, without thinking even when explicitly told to stop.

   when it finally stops to thinking, more then half of the "thinking" are simple anthropomorphic related text.

   … more

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   1 comment  ·  Changes to existing features  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

8. Feature Request: “Auto‑Websearch” / “Always Enable Websearch” in the Personalization Tab

   Hey Proton team,

   thanks for the awesome Lumo 1.2 update! The new personalization options are super handy. To make Lumo’s answers truly real‑time, could you add a toggle under Personalization called “auto‑websearch” (or “always enable websearch”)?

   When turned on, Lumo would automatically pull in up‑to‑date info for any query that needs fresh data—no extra steps required. This would be a game‑changer for users who rely on accurate, real‑time responses.

   Appreciate the consideration!

   11 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   7 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

9. Automatically save chat responses to a document

   In my view, Lumo 2.0 should prioritize its core strengths—leveraging the scientific and technical expertise behind Proton's mission—rather than focusing heavily on image generation. Image creation requires significant financial investment, and many established competitors already serve this market extensively. While I understand there may be legitimate privacy concerns around processing images, it seems the choice to pursue image generation was driven by market trends in AI rather than staying true to the scientific and educational purpose for which Proton was originally founded. My suggestion would be to refocus development resources toward those foundational goals where Proton has unique value and differentiation. For example;

   Background:
   I am deciding at this time who to invest my time, information and money for ai services. I see the benefits of Claude and its ability of Coworker.
   What’s my issue.
   I want to use Lumo for my personal & business for knowledge management with Projects and Drive integration. Currently:
   index works beautifully, but saving responses requires manual copy/paste/upload.

   Requested Feature:
   Automatically save chat responses to a documents name it and to assign to Project folder I’m working in, into the Proton Drive,
   then index those files for future context retrieval without loading full conversation history.

   Thank you for your Suite, I have subscribed for mainly the vpn but have unlimited subscription. I don’t use extensively because the offering is still a bit raw. No sheet app for example or document app. So just waiting.

   Please keep advancing.

   All the best
   Andy

   In my view, Lumo 2.0 should prioritize its core strengths—leveraging the scientific and technical expertise behind Proton's mission—rather than focusing heavily on image generation. Image creation requires significant financial investment, and many established competitors already serve this market extensively. While I understand there may be legitimate privacy concerns around processing images, it seems the choice to pursue image generation was driven by market trends in AI rather than staying true to the scientific and educational purpose for which Proton was originally founded. My suggestion would be to refocus development resources toward those foundational goals where Proton has unique value and… more

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

10. Language correction button for Lumo with categories

    Add a "Report Language Error" button in the Lumo chat interface that lets users flag language issues directly. After clicking, users choose a category:

    Strict — Pure nonsense/incorrect term. Generates a corrected, academically accurate response.

    Synonym — Suggest an alternative word/phrase and context where it fits better. Helps Lumo understand nuance, register and topic-specific synonyms.

    Humoristic — Witty alternatives and wordplay. Helps Lumo learn humor and puns in the local language (especially useful for users with a "Casual/Cheeky" Lumo personality).

    This feeds into Proton's language improvement pipeline, helping Lumo and Scribe produce more natural, localized output — especially for smaller languages like Norwegian, Danish, Swedish and Finish.

    Use case: When Lumo says "hundkur" instead of "hundkoja," the user flags it as Strict. When Lumo misses a pun opportunity, the user flags it as Humoristic — and Lumo gets funnier over time.

    Add a "Report Language Error" button in the Lumo chat interface that lets users flag language issues directly. After clicking, users choose a category:

    Strict — Pure nonsense/incorrect term. Generates a corrected, academically accurate response.

    Synonym — Suggest an alternative word/phrase and context where it fits better. Helps Lumo understand nuance, register and topic-specific synonyms.

    Humoristic — Witty alternatives and wordplay. Helps Lumo learn humor and puns in the local language (especially useful for users with a "Casual/Cheeky" Lumo personality).

    This feeds into Proton's language improvement pipeline, helping Lumo and Scribe produce more natural, localized output — especially for smaller… more

    3 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

11. Abandon Lumo and use resources to address deficiencies in Proton's core secure IT services

    Of course, I have no idea how many people use Lumo or how valuable a business this is to Proton. I hope this suggestion may flush that knowledge out.

    Customers buy Proton services for its secure IT communications. That is - as far as I know - Proton's core business and its unique selling point.

    Many companies offer AI chatbots which - if they're ever useful at all (about which I have doubts) - improve through scale usage, something Proton seems unlikely to achieve, at least when compared to US behemoths.

    Proton appears to have a large development backlog in its core security offerings, if the list in this Proton Voice website is to be believed. Some core deficiencies have remained unaddressed for years. I assume (but don't know) that Proton, as well as other secure communications companies, are trying to deal with a flood of customers fleeing US services or who are nervous about a deteriorating security environment. If so, the demand will only increase.

    Proton makes secure IT better than it makes AI chatbots. Proton's reputation is in secure IT. I assume that Proton sells more secure IT services that AI chatbots. I suggest that AI chatbots are a lower priority for new and existing Proton customers than secure IT. I submit that the world does not need - and Proton is unlikely to sell - more AI chatbots, but the world does need - and Proton can sell - better secure IT services.

    Of course, I have no idea how many people use Lumo or how valuable a business this is to Proton. I hope this suggestion may flush that knowledge out.

    Customers buy Proton services for its secure IT communications. That is - as far as I know - Proton's core business and its unique selling point.

    Many companies offer AI chatbots which - if they're ever useful at all (about which I have doubts) - improve through scale usage, something Proton seems unlikely to achieve, at least when compared to US behemoths.

    Proton appears to have a large development backlog in… more

    2 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    1 comment  ·  Changes to existing features  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

12. lumo extended files

    Subject: Lumo Feature Request: Support for .nix, .zip, and .tar.gz files 🚀Hi Proton Team! 💜First of all, thank you for building such amazing, privacy-focused tools. Lumo looks great and works incredibly well!I would love to share a small feature request that would highly benefit technical users like myself. It would be amazing if Lumo could natively view and handle the following formats:.nix configuration files.zip folders and archives.tar.gz archivesAdding support for these formats would make checking files much faster, eliminating the need to download or move them to external apps just to see what is inside.Thank you so much for your hard work and continuous improvements. Keep up the fantastic job!Best regards,c

    Subject: Lumo Feature Request: Support for .nix, .zip, and .tar.gz files 🚀Hi Proton Team! 💜First of all, thank you for building such amazing, privacy-focused tools. Lumo looks great and works incredibly well!I would love to share a small feature request that would highly benefit technical users like myself. It would be amazing if Lumo could natively view and handle the following formats:.nix configuration files.zip folders and archives.tar.gz archivesAdding support for these formats would make checking files much faster, eliminating the need to download or move them to external apps just to see what is inside.Thank you so much for your… more

    3 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

13. iOS app swipe side bar on left

    I’d like to be able to swipe the left sidebar with my finger in the iOS app both in and out.

    Right now you have to tap the sidebar toggle.

    3 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

14. Allow anonymous data sharing

    I know that Proton doesn't create/train its own AI model, but a good european alternative would be nice to see. As unpleasant as this might be, to create a good model, lots of data is needed. A great app is worth nothing without good models. The current market leader, openAI, uses user data to train their models and have a huge advantage.

    If my data is ever used for training, I want to give it to a company I trust (or think I can trust) to handle them carefully. And this is you. I would like to see the option to anonymously share the chats to help train/refine models. In the best case, only chats without inputs that would give away who asks would be used (this would need to be checked somehow).

    I know that this is probably an unpopular opinion, but we have to face the reality: If we want good open models, we need to provide our data.

    I know that Proton doesn't create/train its own AI model, but a good european alternative would be nice to see. As unpleasant as this might be, to create a good model, lots of data is needed. A great app is worth nothing without good models. The current market leader, openAI, uses user data to train their models and have a huge advantage.

    If my data is ever used for training, I want to give it to a company I trust (or think I can trust) to handle them carefully. And this is you. I would like to see the option… more

    19 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    1 comment  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

15. To make Lumo more human

    Subject: Proposal: A "Dual-Loop" Architecture for Next-Gen LLMs Inspired by Neuro-Mechanical Models

    Dear Research and Engineering Team,

    I am writing to share a conceptual proposal for a new architectural approach to Large Language Models (LLMs), inspired by the neuro-mechanical theories of Lee Kent Hempfling (specifically the distinction between Short-Term and Long-Term memory loops).

    While current LLMs excel at pattern recognition and statistical prediction (effectively acting as massive Long-Term Memory systems), they lack a mechanism for inhibition, self-correction, and proactive intent. This proposal suggests a "Dual-Loop" architecture that could significantly reduce hallucinations, bias, and reactive behaviour.
    The Core Concept: Dual-Loop Architecture

    Current models rely on a single stream of attention mechanisms. I propose a system with two distinct, interacting loops:

    The Long-Term Loop (The Library):
        Function: Standard Transformer weights for retrieving facts, grammar, and historical patterns.
        Role: Fast, reactive, and high-volume data retrieval.
    The Short-Term Loop (The Operator):
        Function: A separate, high-speed module running in parallel.
        Role: Inhibition and Monitoring. Before the Long-Term loop commits to a token, the Short-Term loop evaluates:
            Is this consistent with the immediate context?
            Is this a hallucination or a bias?
            Does this align with the user's intent?
        Mechanism: If a conflict is detected, the Short-Term loop inhibits the output and triggers a re-evaluation cycle (simulating the human "brake" or "pause").
    

    Key Proposed Features

    Dynamic Inhibition Layers: Instead of just predicting the next token, the model actively suppresses low-probability or biased outputs that fail a "truth/intent" check.
    Frequency-Based Temporal Processing: Implementing a "Biological Clock" simulation where different parts of the network process information at different speeds (Fast for "Now"/Logic, Slow for Deep Memory), creating a simulated sense of time and continuity.
    Intent-Driven Generation: Shifting from "Pattern Completion" to "Goal Achievement," where the Short-Term loop sets a constraint/goal before the Long-Term loop retrieves data.
    Modality-Aware Routing: Detecting if a query requires visual simulation or logical deduction and routing processing accordingly, similar to how humans switch between visual and aural dominance.
    

    Potential Benefits

    Reduced Hallucinations: By inhibiting outputs that don't pass a consistency check.
    Lower Bias: Active suppression of training-data biases via the Short-Term "brake."
    Improved Alignment: Better adherence to user intent rather than just statistical likelihood.
    Organic Output: Moving away from rigid, "stiff" text generation toward fluid, "pro-active" responses.
    

    Next Steps

    I believe this approach moves beyond simple scaling (adding more parameters) to architectural innovation. It treats the LLM not just as a database, but as a system capable of self-regulation.

    I would be happy to provide further details, diagrams, or specific mathematical formulations for the "Inhibition Layer" if the team finds this direction promising.

    This Idea was created by Lumo with prompts from me. I thought it was good enough for others to have a better look and did'nt know who to send it to.

    Subject: Proposal: A "Dual-Loop" Architecture for Next-Gen LLMs Inspired by Neuro-Mechanical Models

    Dear Research and Engineering Team,

    I am writing to share a conceptual proposal for a new architectural approach to Large Language Models (LLMs), inspired by the neuro-mechanical theories of Lee Kent Hempfling (specifically the distinction between Short-Term and Long-Term memory loops).

    While current LLMs excel at pattern recognition and statistical prediction (effectively acting as massive Long-Term Memory systems), they lack a mechanism for inhibition, self-correction, and proactive intent. This proposal suggests a "Dual-Loop" architecture that could significantly reduce hallucinations, bias, and reactive behaviour.
    The Core Concept: Dual-Loop Architecture

    … more

    3 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    2 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

16. There should be a way to turn off the cat avatar

    I would love to use Lumo, but can not because of the annoying cat avatar. Lumo says it could be turned off, but gives misleading instructions how to do it. For example the Settings menus are very different what Lumo desribes and what I actually see.

    8 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    2 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

17. Create doc and spreadsheet directly from chat

    Add the ability to automatically open a chat window directly within a document and embed a table in Proton Sheets. This would streamline real‑time collaboration by allowing users to start a conversation or view a related spreadsheet without leaving the current file.

    8 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

18. Add support for de-googled devices

    Lumo is unsupported without Google Play Services. Can't start the application, blocked during boot.
    Using LineageOS. Should work without Google.

    22 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

19. Using better open-source/weight AI Models

    Hi everyone,

    I am writing this because I think Lumo has so much more potential. Imagine them using a 1 trillion parameter Chinese open-source/weight AI model like it would give them so much more flexibility in new features and accuracy of up-to-date data. They could even train an own AI model that they can then use like OpenAI uses ChatGPT as an multimodal AI model without having to rely on others creating open-source/weight AI models for them to use plus they could customize it perfectly for privacy, speed and accuracy. This own AI model could be based of some other open-source/weight AI model or could be designed specifically for Protons privacy focused servers. They could even integrate an own search engine into this customized model, enabling it to use auto websearch so you don’t have to actively turn it on in every new conversation and could be trained for not only text but also image generation and upload, video generation and upload, text to speech, speech to text and even a Live mode like Gemini, ChatGPT and Perplexity.

    Hi everyone,

    I am writing this because I think Lumo has so much more potential. Imagine them using a 1 trillion parameter Chinese open-source/weight AI model like it would give them so much more flexibility in new features and accuracy of up-to-date data. They could even train an own AI model that they can then use like OpenAI uses ChatGPT as an multimodal AI model without having to rely on others creating open-source/weight AI models for them to use plus they could customize it perfectly for privacy, speed and accuracy. This own AI model could be based of some other… more

    13 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

20. RAG my proton drive. Make Lumo be able to chat with me about my files.

    I bought into Lumo and upgraded my proton drive as I had the impression that I could query my files.
    In fact, I can only upload some of them. Well, some, in fact only 10Mb before it complains. Then if I leave the chat and come back, it seems to have forgotten I uploaded a file to chat with.

    Here is my use case which is currently available in CoPilot.
    1. Have an option for Lumo to chunk and ingest the files in my drive.
    2. Allow me to query them like a proper RAG system

    Thanks.

    9 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating