De-anonymization attacks against the privacy coin XMR

The USA in specific has had a similar problem before with encryption being classified as a munition making very problematic to import or export encryption. That's actually pretty well documented in various pieces of Java code from Sun if you're curious, because different algorithms could not be part of the JRE/JDK that was distributed publicly.

Your mention of "highly centralized billion dollar compute operations" is actually related to the training of the models not the inference. Doing inference for many of these models is readily available at modest consumer hardware availability. There are many different ways to break up models (MoE) etc. The notion that you need a large super computer to do inference is unfounded.

Also, as a reminder, cryptocurrency mining has already proven this to be a thing. Some stay above ground, some go to geopolitical areas for shelter and some stay underground entirely.

For your entertainment I will also include a more simplified play-by-play of how this can play out in the near future:

1. OpenAI or some other USA based AI company continues to get outplayed by foreign models (Qwen, Deepseek)

2. Company cries to government

3. Government does a similar munitions or tariff to what we saw with encryption. Requires at the least anyone wanting to use AI gets one from the good boy list etc.

Now you either (a) use only AI from the good boy list and get outplayed in the global marketplace where our main export is global technology or (b) start acting like a Chinese citizen and using a VPN to access AI services not available only on the approved good boy list.

I will stop here because the rest is already very well documented with how this progresses and you get the same result as the darknet marketplaces (DNM). DNS censorship for AI services not on the good boy list. DNS censorship and legal pressure for VPNs that allow non-good-boy-list services, etc.

Why wouldn't someone change a few .com endpoints to .onion and keep it moving while you send some coin to a wallet?

Unfortunately, it doesn't work like that. The EAE attacks only require that the end destination is colluding with the start destination.

Like everything with decoys, privacy is stochastic. So I wouldn't go around making absolute claims about the privacy as many proponents of monero like to do. The developers advise against making these sorts of claims. Monero makes privacy a lot easier, but it's not perfect.

>There are court cases that give the confidence necessary. It is also something to stay abreast of. Always just ask yourself who the transaction is intended to be hidden from.

In the free world, we have the concept of innocent-until-proven-guilty and evidence-beyond-a-reasonable-doubt. Decoy-based approaches give you plausible deniability, but this often isn't enough for more domains where a lower standard of proof is needed.

Fortunately, all this and more will be fixed in FCMP++ upgrade.

Once you create a wallet and write down the seed phrase, generate a "view key". Creating a new wallet from this "view key" allows you to see incoming transactions to your addresses, but not spend them. So you don't need as much security for "view-only" wallets.

You can generate an address from either wallet. It's a long string of numbers and letters that begins with an "8", under "Receive".

MWEB is certainly an improvement over transparent transactions (and other methods such as coinjoin, coinswap, cashfusion, etc.), and I welcome the litecoin upgrade. I agree that decoy-based privacy is weak.

However, I don't believe that the mimblewimble meets the standard of privacy needed for most users. It's not the visible amounts and addresses, but the links between transactions that are the main problem. CTs on their own are just a "nice-to-have".

The end goal should be a zcash or firo style of privacy. I think you can scale that to a global network with an adjustable block size, payment channels, and atomic swaps between multiple cryptocurrencies. The problem is that zcash and firo have weak tokenomics compared to monero. Grin will have a hard time finding an initial niche that isn't currently satisfied by monero, and if it does take off, its changes could be merged into bitcoin (https://www.truthcoin.info/blog/imex/).