Yubico在生产YubiKey的时候会在插槽1放入一个OTP凭据，Yubico提供的凭据ID开头是“cc”，后期上传的凭据ID开头是“”。Yubico提供的OTP凭据是受信任的，后期上传的没有出厂自带的那么可信，因为在传输和保存的时候私钥可能会泄露。这是官方的说法：’‘ prefix credentials are not guaranteed to have the same availability as production ‘cc’ prefix credentials. Yubico reserves the right to revoke any ‘‘ prefix credential on the Yubico validation service (YubiCloud) at any time, for any reason, including if abuse is detected or if the credential is loaded onto a counterfeit YubiKey. (From Yubico AES Key Upload)
The tremendous development of artificial intelligence technology in recent years has made this technology famous, people even can hear this word from a primary school student who just knows how to type. However, as an important part of artificial intelligence, “machine learning” is not so famous. However, machine learning is a shortcut to “intimate contact” between ordinary people and artificial intelligence, even if they do not have in-depth study in computer science.
So, what is machine learning? You can simply understand from the literal meaning: give computers the ability to “learn”. Machine learning is a very interesting technique. It uses technical means to enable computers to analyze the information they receive and to form an “understanding” of the information. Machine learning is a very esoteric technique, but it is friendlier to the average user than to the developer. In other words, if people don’t plan to write a machine learning algorithm and just apply machine learning, it will be very simple and attractive.
In order to make simple use of existing machine learning techniques, we can use some open source libraries to achieve. For example, for most students, writing a TOEFL essay is a painful thing, and if we look through common machine learning documents, we will find a library called “scikit” for literary analysis such as semantic recognition. Therefore, we only need to write a program that uses this library (this is still difficult, but obviously it’s easier than writing a machine learning algorithm from the beginning, and it seems to have been written by some machine learning experts). So, we just import a few TOEFL essays we wrote before (this is called “training”), and then import a sample of the TOEFL essay you are about to write, the computer can analyze your logic, style and other writing habits and the sample. Then, rewrite “your own” TOEFL essay, does this sound great? (I am not trying to provide a lazy way to write a TOEFL essay, it is important to learn English well) Of course, machine learning can do a lot of other things like face recognition, OCR, even Alpha Go uses machine learning. So, machine learning fascinates me because machine learning can provide a quick way for ordinary people to solve practical problems.