Software, protocol, hardware, browser extension, service model, anything. The bar is that it doesn’t really exist yet, or only exists in a form that fails the threat model. What problem are you still solving with duct tape?
Useful answers look like: the concrete leak or friction, why current options fall short, and the rough shape of the thing that would fix it.
I want a private file share 2-5 GB. No login. Monero payment. Fixed host period.
I've tried this https://file.svoboda.center/
many others. nothing is simple.
https://wormhole.app/ and it's free.
amazing but my why is it free radar is popping off
A tool that actually removes your face from all online photos and also removes personal information. The current process is tedious and incomplete without direct effort even if you hire services to do it.
This may exist, but an easy way to provide false facial recognition data for things that require it. I don't want to use my real biometrics.
Something to foil gait recognition without making walking super awkward would be nice too.
A tool to anonymize writing style would be cool too then someone could write radical stuff using a pseudonym and still be able to write regular books, articles, etc. without fear of discovery.
A tool that provides poisonous data to invasive services when they try to take it without consent would be nice too.
Also, a tool that defeats ALPRs, but doesn't interfere with a cops physical line of sight human plate verification. Preferable an easily hidden one.
This is called adversarial stylometry, and partial tools exist. BTW, a writing-style tool / LLM alone does not make radical pseudonymous writing unlinkable from your normal books or posts if the samples are long. Operational separation (different devices, accounts, timing, and topics) matters as much as the prose. Treat any rewriter as a reduction in risk, not a guarantee: topics, arguments, private facts, timestamps, etc can still connect a pseudonym to signed work.
You might like this fashion too lol (jk - tbh it’s pretty useless, but it looks funny, and it lets everyone know you’re trying to avoid tracking):
https://urban-privacy.com
https://www.capable.design/
I actually have one I think is genuinely better than my initial suggestions:
Provably private ancestry and provably private genetics. I want info about my DNA and ancestry without being forced to produce that in a relevant way to any others to include Nation State Actors. Is it reasonable to expect me to not be able to get genetic screening or ancestral history because I give a fuck about privacy?!
Your main goal is to keep your DNA sample/results unlinked from any government-issued ID (same reason I buy hardware with cash so the purchase isn’t tied to a bank account -> govt-ID).
People do exactly this. It’s extra annoying steps, but it works in practice for decoupling the test from your legal ID:
Buy the kit in person at a pharmacy/big-box store with cash (or use an Amazon gift card bought with cash / prepaid debit / proxy-shopper).
Register the kit with a completely fake name + throwaway/burner email.
Pay any lab activation fee with a cash-bought prepaid debit card / proxy-shopper.
Use a VPN when activating/viewing results.
Opt out of relative matching and research sharing.
Labs mostly just track the sample by barcode, not by verified legal identity, so the above breaks the obvious payment/shipping/account trail back to your gov ID.
BUT(!): DNA itself is a unique biometric, so relatives who test (or future law-enforcement uploads to open databases) can still re-identify you in some cases. Companies can also be compelled by court order or suffer breaches.
I’d never do DNA for that^^ reason.
There is a startup (with an AI-looking website = not trustworthy IMO) that claims to be doing exactly the local/air-gapped approach: https://privdna.com/
And if you are really determined you can do it fully locally now too:
https://themenonlab.blog/blog/home-genome-sequencing-guide (https://iwantosequencemygenomeathome.com/ - full DIY home sequencing, true local, no third party, nice weekend project lol :)))
I can think of many problems but here's a cut of just three I was arguing about.
Most OPSEC spaces recommend average Linux distributions as private and secure operating systems when they are not really doing anything leading in either front. Many also move security backwards, like Debian and derivatives freezing updates for applications. It ends up with these choices being decided by people saying "It is not Windows or Mac" rather than saying "It is more private and secure than the other [distributions]".
Secureblue is the closest I think exists for Linux right now, but it is quite different from what most Linux users would expect from daily use experience and also novel. I think if we're going all-in then OSes would leave Linux entirely.
I am aware Qubes OS exists but Qubes OS is simply a hypervisor running, again, average Linux distributions as virtual machines. Create an operating system designed around privacy/security first and you'll add extra defences to Qubes OS through more secure virtual machines as well.
This more secure distribution would then downstream into allowing more secure XMR nodes to be hosted, web servers, other services...
It's a really personal thing but I think we can make something better than Thunderbird that is just as functional while also being much smaller. I want to have a low attack surface email client. A lot of other competitive email clients don't have the same scrutiny and usage and are made with languages I don't like.
I think with the increasing attacks on individuals known to possess hardware signers, I think it is good to work harder on having secure wallets in devices that appear less suspicious, like phones. They're in no way made to replace cold storage, but alleviate risks.
I say 'cryptocurrency' because many are guilty of this, but you can think XMR, BTC, whatever here.
A lot of wallets that have leading cryptocurrency feature support are often made with weird language/stack decisions I hate. They make great references for the cryptocurrency's features but they do not make good wallets or mobile apps. They can have poor performance, UX and worse lack distinct security and privacy features.
I like Cake Wallet, but I don't like it enough. It feels odd to use at times, this is solely a personal preference. ZEUS is a leading Lightning wallet/node app but it's made in TypeScript and I find it very slow to use compared to the custodial Phoenix Wallet which is made with native Android libraries and Kotlin and is fast.
Developer stagnation is an issue, Samourai Wallet was mostly the same legacy Java codebase with years for major features to come, and I don't see much changes in Ashigaru either. A lot of XMR-only wallet apps are stagnant with exception of Monfluo which has several improvements open in the issue tracker.
For Android/GrapheneOS, if someone made a single or multi-wallet with Kotlin, mostly native libraries and UI, it would make a killing. Would also ensure compatibility with various exploit protections in GrapheneOS too.
My killer checklist would be:
a cash-to-monero machine that does not want your face, phone, or a selfie. the ones that exist still want an id or a camera.
and a way to prove you are old enough without proving who you are. age check is the new id grab.
deleted by author