Δημοσίευση
Mr AKRAPOKAM_77
·
--
TURNING MACHINE INTELLIGENCE INTO VERIFIED KNOWLEDGE.
Brother, machine intelligence has become so powerful these days that it seems to know everything, but the real problem is that sometimes it confidently says absolutely nonsense or even lies. Meaning, whatever seems right to hear, turns out to be wrong when checked. This is why doctors, lawyers or people in finance cannot trust it completely. Here a new system named Mira Network has come up which works in a completely different way. It checks the content coming out of the machine thoroughly and sends only the confirmed things forward.
The idea is simple, take any text report code or anything and break it into small pieces containing truth. For example, if there are two things in one line, then make separate claims for both. Then send these small pieces to many different checking machines. Each machine sees according to its own calculation whether it is right or wrong. When many machines say yes to the same thing, then a digital certificate is created which says yes brother, this is definitely correct.
The work goes like this, you or any company puts its content. It tells how strict checking is required, more strict for medical things and a little less for normal things. The network divides that content into pieces and then randomly sends it to many computer nodes. Every node runs its own model and answers yes or no or chooses from the options. All the answers are counted and when a large majority agrees then the certificate is ready. By showing this certificate, you can say that this thing is verified.
Full attention is also given to privacy, the entire content is not shown at one place, everything is distributed randomly. For security, nodes deposit money as stake, whoever checks correctly will be rewarded, if he does it wrong then the money will be deducted. That is why most of the people do it right, doing it wrong is costly.
This network has its own token MIRA. Fees are paid, staked, voted and rewarded. As more people use it, the fees increase. As more people join, checking becomes better. Something with 70 percent accuracy from one machine goes above 95 percent because many different models check together.
The entire system is decentralized and no single company or person controls it. Initially, some checking is done and later everything is done randomly so that no group can cheat. It is such that doing the right thing is in everyone's interest.
The advantage is that now trust in machine intelligence can increase. Reports can be checked in hospital, documents can be verified in law, data can be confirmed in finance without a human being sitting and watching every time. Whether the content is written by a machine or a human, it works for both. In future,may become a big database in which only verified facts can be stored on the blockchain.
To be honest, after seeing things like Mira, it seems that in future machine intelligence can become truly reliable without the fear of lying.
#Mira $MIRA @mira_network
The idea is simple, take any text report code or anything and break it into small pieces containing truth. For example, if there are two things in one line, then make separate claims for both. Then send these small pieces to many different checking machines. Each machine sees according to its own calculation whether it is right or wrong. When many machines say yes to the same thing, then a digital certificate is created which says yes brother, this is definitely correct.
The work goes like this, you or any company puts its content. It tells how strict checking is required, more strict for medical things and a little less for normal things. The network divides that content into pieces and then randomly sends it to many computer nodes. Every node runs its own model and answers yes or no or chooses from the options. All the answers are counted and when a large majority agrees then the certificate is ready. By showing this certificate, you can say that this thing is verified.
Full attention is also given to privacy, the entire content is not shown at one place, everything is distributed randomly. For security, nodes deposit money as stake, whoever checks correctly will be rewarded, if he does it wrong then the money will be deducted. That is why most of the people do it right, doing it wrong is costly.
This network has its own token MIRA. Fees are paid, staked, voted and rewarded. As more people use it, the fees increase. As more people join, checking becomes better. Something with 70 percent accuracy from one machine goes above 95 percent because many different models check together.
The entire system is decentralized and no single company or person controls it. Initially, some checking is done and later everything is done randomly so that no group can cheat. It is such that doing the right thing is in everyone's interest.
The advantage is that now trust in machine intelligence can increase. Reports can be checked in hospital, documents can be verified in law, data can be confirmed in finance without a human being sitting and watching every time. Whether the content is written by a machine or a human, it works for both. In future,may become a big database in which only verified facts can be stored on the blockchain.
To be honest, after seeing things like Mira, it seems that in future machine intelligence can become truly reliable without the fear of lying.
#Mira $MIRA @mira_network
Αποποίηση ευθυνών: Περιλαμβάνει γνώμες τρίτων. Δεν είναι οικονομική συμβουλή. Ενδέχεται να περιλαμβάνει χορηγούμενο περιεχόμενο. Δείτε τους Όρους και προϋποθέσεις.