Blockchain analysis tools are being used by the IRS, so this is already happening. Albeit at a relatively small, yet targeted scale:
http://fortune.com/2017/08/22/irs-tax-cheats-bitcoin-chainalysis/
I'm not sure whether Chainalysis uses neural networks though.
It will be interesting to see what the impact of 2nd layer scaling solutions on blockchain analysis will be, given that a lot of the data that was previously publicly available will then be missing.
Topic: Data Science In Blockchain Technology: cnn-proof Blockchains (Read 160 times)
Jeff Garzik is working on "bloq", a company that does exactly that (to analyze transactions and try to guess where the money is coming from). This is one of the main reasons I never trusted Jeff Garzik, I think he sold out for the NSA or something.
As far as if this is viable or not, im not sure. I tested back then a mixer, I think it was helix, and it had so many options, like mixing the coins with freshly minted coins and then recieve back in random batches at random times and through Tor. Trying to trace back the original transaction would be a mess and a waste of resources.
Now I regret doing that test, because I did it with like 1 BTC, and now that 1 BTC I cannot sell, since the government would ask me where that money came from, and explaining to them the truth (that I was just testing a mixer) might not cut it and they may just steal my money.
As far as if this is viable or not, im not sure. I tested back then a mixer, I think it was helix, and it had so many options, like mixing the coins with freshly minted coins and then recieve back in random batches at random times and through Tor. Trying to trace back the original transaction would be a mess and a waste of resources.
Now I regret doing that test, because I did it with like 1 BTC, and now that 1 BTC I cannot sell, since the government would ask me where that money came from, and explaining to them the truth (that I was just testing a mixer) might not cut it and they may just steal my money.
I lately thought much about the growth of data science and blockchain technology.
Both sectors are making huge progresses at the time we are speaking. Yet, both technologies currently grow separately.
Todays Convolution Neuronal Networks have already exceeded human sensory capabilites. Their "weak" point is the lack of data in many fields, that prevents a sufficent training for optimal pattern recognition.
If something doesn't lack in blockchain technology, then it's data. The data is also free to access and distinct.
Thus, I expect an expansion of pattern recognition systems into the field of blockchain technology.
In the current stage of cryptocurrency acceptance this shouldn't be a threat. But in the near future, when cryptocurrency spread its usage and develope more use cases, pattern recognition becomes a powerful tool to analyze the blockchain and its users. Although wallet addresses are not bound to an actual indivdiual, its patterns can draw conclusions about personal behaviour, preferences, circumstances. Only small amounts of meta data are able to allow determined inference, which carries huge power with it.
Like most technological enhancements this implies many benefits, but it might also imply dangers, especially if this data gets not treated responsibly.
Many governments are working on software to analyze blockchains. Not in terms of pattern recoginition, but to be able evaluate specific cash flows. This aspect becomes important when it comes to tax payments.
Nevertheless I am very confident that it's only a matter of time until cnn's absorb blockchains. That's why it's in my opinion important to discuss the consequences and how cryptocurrencies should handle this matter. Developing CNN-proof blockchains could get a thing in the future and become major feature for privacy coins.
Why are your opinions?
Both sectors are making huge progresses at the time we are speaking. Yet, both technologies currently grow separately.
Todays Convolution Neuronal Networks have already exceeded human sensory capabilites. Their "weak" point is the lack of data in many fields, that prevents a sufficent training for optimal pattern recognition.
If something doesn't lack in blockchain technology, then it's data. The data is also free to access and distinct.
Thus, I expect an expansion of pattern recognition systems into the field of blockchain technology.
In the current stage of cryptocurrency acceptance this shouldn't be a threat. But in the near future, when cryptocurrency spread its usage and develope more use cases, pattern recognition becomes a powerful tool to analyze the blockchain and its users. Although wallet addresses are not bound to an actual indivdiual, its patterns can draw conclusions about personal behaviour, preferences, circumstances. Only small amounts of meta data are able to allow determined inference, which carries huge power with it.
Like most technological enhancements this implies many benefits, but it might also imply dangers, especially if this data gets not treated responsibly.
Many governments are working on software to analyze blockchains. Not in terms of pattern recoginition, but to be able evaluate specific cash flows. This aspect becomes important when it comes to tax payments.
Nevertheless I am very confident that it's only a matter of time until cnn's absorb blockchains. That's why it's in my opinion important to discuss the consequences and how cryptocurrencies should handle this matter. Developing CNN-proof blockchains could get a thing in the future and become major feature for privacy coins.
Why are your opinions?
Jump to: