🚀 Maple AI Launch! Our end-to-end encrypted AI chat app, Maple AI, is now available! Try it for free and experience private AI conversations that sync automatically in private across your devices.
https://trymaple.ai
48% of organizations and businesses risk security and intellectual property by entering non-public info into AI apps. Maple AI changes the game with end-to-end encryption for secure chats. Work and personal use, it's all private between you and the LLM.
Our server code is open source, so you can verify the security and integrity of our platform. Confidential computing ensures your conversations remain private, with cryptographic proof based on reproducible builds. Don't trust us to be honest, verify.
We offer flexible pricing plans: Free (10 chats/week), Starter ($5.99/month), and Pro ($20/month). Pay with credit card or Bitcoin. Prices are discounted until Jan 31st!
We wrote up more explaining the launch on our blog:
https://blog.opensecret.cloud/maple-private-ai-for-work-and-personal
Are you a developer? Maple AI is the first app built on our revolutionary new platform, which provides a suite of secure tools for developers. With OpenSecret, you can add secure auth, private key management, encrypted data sync, private AI, and more to your own app.
https://blog.opensecret.cloud/introducing-opensecret/
Sign up for Maple AI now and experience the future of secure and private AI. Our open-source technology and end-to-end encryption keep your conversations safe and private, on all your devices.
https://trymaple.ai
Replies (37)
Will check it out.
Is there a plan for how to implement search to find past conversations? Is that possible?
I've seen similar claims before but my distrust hinges on my distrust for "Confidential computing" or other "secure element" type of hardware that promise to be unhackable.
Everybody telling a hand full of LLMs their biggest problems in work and life is a real problem but those most interested in that data would be the most incentivized to launch something like "Maple AI" and how would you proof that you're not a CIA cutout? Cause state actors clearly have backdoors to such devices that are not publicly known.
👀
nostr:nevent1qvzqqqqqqypzphchxfm3ste32hfhkvczzxapme9gz5qvqtget6tylyd7wa8vjecgqyshwumn8ghj7en9v4j8xtnwdaehgu3wvfskuep0w3uhqetnvdexjur5qyfhwumn8ghj7ur4wfcxcetsv9njuetn9uq32amnwvaz7tmjv4kxz7fwv3sk6atn9e5k7tcqyp0809cmss35wtkep834ekvq8gwc5tqxjf95uvzn5mnpkq7dzltq7ugmcj6
👀
💯
> 48% of organizations and businesses risk security and intellectual property by entering non-public info into AI apps. Maple AI changes the game with end-to-end encryption for secure chats. Work and personal use, it's all private between you and the LLM.
E2EE isn’t enough. Businesses risk their data because of the organization controlling the LLM on the other end of the connection. All encryption does is ensure the data isn’t intercepted on the way.
Can you talk a little bit about what your organization does differently once the data reaches the AI chatbot?
Nostr?
This is a new level of encryption. Confidential computing uses enclaves on both the CPU and the GPU to encrypt the data while processing. The services use attestation to verify that they are running the proper software and sever the connection if the code has been tampered with.
So yes, E2EE typically is about encrypting data in transit. Now we can encrypt on the processing unit and in the memory.
Just bought an annual plan for Maple with bitcoin.
"It's all private between you and the LLM."
nostr:nevent1qvzqqqqqqypzphchxfm3ste32hfhkvczzxapme9gz5qvqtget6tylyd7wa8vjecgqyshwumn8ghj7en9v4j8xtnwdaehgu3wvfskuep0w3uhqetnvdexjur5qyfhwumn8ghj7ur4wfcxcetsv9njuetn9uqzqhnhjudcgg689mvsnc6umxqr58v29srfyj6wxpf6desmq0x304s0hsjpxv
You still share user data and analytics with your partners.
How about don't?
went to upgrade, but only got a pay with credit card option. did I miss the pay with btc option? also, any chance of giving an annual subscription option?
yup i totally missed it...let's make that option larger please 😑
annual plan acquired, thanks!
Signed up for a yearly plan, just to support the project. My use cases for LLMs are very particular so here's some things I hope to see in the future:
1. Nostr login support
2. File upload support for data analysis (pdfs, etc). Integration with a Blossom server a plus.
3. Image generation (preferably with the ability to save to a Blossom server lol)
4. Web search. LLMs are great for analyzing data and providing context, but they are even better when they can crawl the web, particularly for specific types of problems.
nostr:nevent1qvzqqqqqqypzphchxfm3ste32hfhkvczzxapme9gz5qvqtget6tylyd7wa8vjecgqyshwumn8ghj7en9v4j8xtnwdaehgu3wvfskuep0w3uhqetnvdexjur5qyfhwumn8ghj7ur4wfcxcetsv9njuetn9uq32amnwvaz7tmjv4kxz7fwv3sk6atn9e5k7tcqyp0809cmss35wtkep834ekvq8gwc5tqxjf95uvzn5mnpkq7dzltq7ugmcj6
Is there an android app or it's only browser based for now ?
Browser for now. You can do "Add to Home Screen" from the browser and use it like an app. That's personally how I use it.
What kind of functionality would you like to see if it were a native app?
Which models can you access when signing up? Do the models have access to the internet?
I would love to be able to choose a model llm christian/conservative where the prompt answers with that world view.
For example, no biological men should ever compete in women's sports. here is what Maple had to say : ' It's complicated '
Open source and you charge money… you claim non-private data is being send to AI APIs and yet you build one yourself?
Who are you?
Private AI inference! You’re just yet another spin up
Contrary to popular belief, enclaves do not have encrypted state
The isolation and confidentiality is enforced by the AWS Nitro hypervisor
Any code on the hypervisor can bypass confidentiality
Is there an iOS app? Android?
As a layperson your comment comes across sounds like a back door. Do you mind expanding?
You always have to trust someone
The end
Tons of companies charge money for open source, are you new?
So the connection is end-to-end encrypted, but can the admin of the server (you or amazon) see any of the requests or data? How am I to verify that is private?
Dumb question. What LLM are you chatting with when using Maple? ChatGPT? Other?
They charge for their own hosting of their open source - yes.
And yes I am new.
But these people are taking about their eNcRyPtEd AI hosting. But like… no wtf. If I need a TRUE private AI solution I have to have my own inference.
You can’t say “company A has your data so use us instead”
The real solution is self-hosting - not trusting yes another company that promises some “encrypted” service. Encryption is ubiquitous, and so is data mining. 🥱
Yes I am new, btw.
So new. The newest.
No one is newer than me
Don't come in hot then on people providing a service that people want just cuz you're not their demo. Do I want to run my own models, yes. But the 1070 I have won't even be able to run anything better than glorified man pages. Do I want to buy a new one, sure, but for now I'll use private services like this over ones I know are collecting my data
You’re sending your data to this company and assuming they’re not collecting it, first off.
I’m not coming in hot for saying that - to a nostr/bitcoin crowd we should all understand “not your keys not your coins” as well as “not your computer not your data.”
They are also advertising their “encryption” ffs. Which AI company DOESN’T use encryption? That’s straight up either bad marketing or they know most people aren’t smart enough to not fall for it.
Go buy a Mac Mini with 32GB of ram and you can run a ton of LLMs at 40+ tokens/second and do image generation
I think you need to scroll back through this accounts posts. They have clout on nostr as Cypherpunks, it's literally a business their wallet was one of the first with nostr integration and was highly regarded, I could continue.
It’s like you just told me to check out a PoS coin… answer is no, by default. PoS is the problem. If people are uncomfortable sending their queries to OpenAI, at al, then using this company solves nothing. I can’t appeal to the masses just because the “nostr” crowd likes them. It changes nothing.
A company telling me to use them because they are private/encrypted (which is unprovable) will always lose to private AI inference - which isn’t expensive to do (less than $500)
I am completely onboard with spinning up an AI company, having people pay in bitcoin/ecash, and having all my users be anonymous. I want to be clear.
I am working on exactly this.
I want my users to be random npubs, pay with cashu and be comfortable with the fact that I can inevitably see all their queries… but don’t know who they are - unless they dox themselves in their query. I have no interest in snooping their shit, but obvi I can’t help but have the ability to see it all.
It is foolish to think an entity you’re sending data to won’t or can’t parse / retain it.
I need my users to deeply understand this. I can’t tell them to not worry cuz I use “encryption.” Just use my service and don’t send anything you can’t tolerate getting out. For secret stuff you should always use truly private LOCAL AI inference
nostr:npub1az9xj85cmxv8e9j9y80lvqp97crsqdu2fpu3srwthd99qfu9qsgstam8y8 has mentioned this problem with his own unleashed.chat.
The 50% discount isn’t available if paying with bitcoin? nostr:npub136jg2fnty2z5vwcnh7p4jpckrs3tk0dpueftgs7mznuuaenjpfps6tjnxf
Can you please share any info about what models are being used?
