Pinned Tweet
Maple Work is here: a private AI agent for your computer.
Give it a task. It reads your files, uses your apps and gets it done, on the latest open models.
Using powerful AI shouldn't mean lowering your security standards. Welcome to the open frontier.
Most AI privacy is a policy: a promise not to look.
Maple's is by design. The model runs in an enclave Maple can't see into, and you can check it yourself.
@BTCsessions shows how.
Don't think I'm crazy but... I just gave AI my tax information and blood tests.
This video on @TryMapleAI will show you why I'm not worried.
Full tutorial on @SOVsessions 👇
youtu.be/4PCt6EEEEU0
Worth the full hour, but if you're short on time:
35:02 why an enclave beats a promise
6:07 checking software before you install it
18:42 a year of statements into a tax package
30:13 lab results into a health dashboard
youtu.be/4PCt6EEEEU0?t=2102
74
Maple AI retweeted
This is really exciting for @buzzdotxyz users who don't want to send their community's data to third parties that can read everything.
Use the Maple agent to connect their harness to powerful open weight models running in a TEE, where nobody can see your data.
Private Maple agents, now in @buzzdotxyz.
Mention one in a channel. It works on your computer and answers in the thread. It runs open-weight models like DeepSeek and Kimi in secure enclaves, so we can't read what your team asks it, or anything it reads to answer.
Maple AI retweeted
Don't think I'm crazy but... I just gave AI my tax information and blood tests.
This video on @TryMapleAI will show you why I'm not worried.
Full tutorial on @SOVsessions 👇
youtu.be/4PCt6EEEEU0
Private Maple agents, now in @buzzdotxyz.
Mention one in a channel. It works on your computer and answers in the thread. It runs open-weight models like DeepSeek and Kimi in secure enclaves, so we can't read what your team asks it, or anything it reads to answer.
Setup:
1. Install Maple Work: maple.ai/work
2. In Buzz: Settings → Plugins → Add, paste MaplePrivacyLabs/buzz-maple-plugin, click Find, then Install
3. Create an agent and pick Maple as its harness
Full guide: github.com/MaplePrivacyLabs/…
2
5
322
You don't have to take our word that we can't read it. Run the same check the Maple app does: maple.ai/proof
2
214
John's right, and most researchers don't notice the trade they're making.
Unpublished data. Grant drafts. Interview transcripts. Peer reviews. All sent to a company that can read them.
Maple Work is an AI agent for your research folders, on open models in enclaves we can't read. You can check.
Fellow researchers: if you're using AI, you should be doing so privately.
Why send your research data & intellectual property to private AI companies?
"no training" & similar promises don't go nearly far enough.
Good news: in ~ 5 minutes, you can be using AI with the ease of ChatGPT but much better privacy.
Easy Mode: AI in Encrypted Enclaves
Today, companies like @TinfoilAI and @TryMapleAI serve open-weight AI models inside encrypted enclaves.
Using these services is a huge level-up: they don't train on your prompts. In fact, they can't even see them. Risk of your prompts getting exposed from a breach goes way down, too.
No risk goes to zero, but this is a big, meaningful improvement. Just a handful of minutes to get started.
Bigger Picture: Bring Inference Back To the University
Universities should be hosting open-weight models, fine-tuning them, and making them dead-easy for researchers to work with.
Ideally, universities are serving this inference in encrypted enclaves & developing research lines on how to do it effectively.
Huge leadership potential & public benefits for academic and intellectual freedom, and open knowledge.
Want to dip your toes in encrypted enclaves? @micahflee has a nice recent piece:
micahflee.com/heres-how-anyo…
Inkling 2 Small is now in Maple.
@thinkymachines' open-weight model is built for long work. In Maple Work, we had it:
→ turn a year of business statements into a P&L and flag what's off
→ build a 3D maple forest from one prompt
Connect Square, Stripe and more. Privately.
Available on all paid plans.
What it's good at, where it's weak, and how to use it privately: maple.ai/blog/inkling-2-smal…
6
222
Inkling 2 Small is now live in Maple as well!
Grateful to keep building with Tinfoil.
Inkling 2 Small from @thinkymachines is now on Tinfoil Chat and Inference API.
Great at agentic coding, vision and knowledge work, especially with spreadsheets and charts.
It runs in a secure enclave, so you can verify what you share stays between you and the model. 🩶
Maple AI retweeted
Pay for private AI with bitcoin
Excellent writeup from @micahflee on how to start using private AI. Talks about Maple Chat and Maple Work. 🔥🔥🔥
@TryMapleAI
micahflee.com/heres-how-anyo…
OnionShare creator Micah Lee (@micahflee) wrote a step-by-step guide to private AI with Maple, from his first chat to handing Maple Work a real job.
His point: you no longer need a rack of GPUs or LLM expertise to use open models privately.
He asked Maple Work whether Bluesky publishes its list of trusted verifiers.
“It did a bunch of research, figured out that it's nontrivial to get the list, and wrote a Python script for me that does it anyway.”
1
3
407
His full walkthrough covers sign-up, private chat and setting up Maple Work on your own computer:
micahflee.com/heres-how-anyo…
1
6
293
Maple AI retweeted
We know that you have data in a bunch of apps and services. Use them securely while thinking and reasoning.
Private AI doesn’t mean you have to live off the grid. Connect to your existing ecosystem.
What cool things are you building this weekend with Maple Work?
By design, we have absolutely no idea 🥲
Maple Work connects to 26+ apps in one click, or any MCP server. Privately.
Pick one and give it a task this weekend.
Your app sign-ins stay on your computer.
What your apps send is encrypted on the way to the model, and Maple can't read it.
Get Maple Work 👇 maple.ai/work
2
272
















