Decentralized data, Artificial intelligence, and Cartesi
Decentralized data and artificial intelligence is a combination of deep learning and blockchain. Without the Https protocol, everything we do with the web will not exist. There is web 1.0 and then there is web 2.0 that is centralized and we give our data for free in exchange for some services. And now we are launched into Web 3.0 which is decentralized, autonomous, and peer to peer. With web 2.0, anyone can use your data, sell your data, profit from your data.
Web 2.0 and Web 3.0
The challenge with this is that all these data are contained in the centralized servers and these servers have single points of failure, as we gradually move into the world of Web 3.0, data will be the most powerful tool. But right now, you are giving it away for free and when you don’t own your data, you cannot deliver value with it.
When robotics, artificial intelligence, deep learning, and automation becomes a complete reality to the average man. One other difference between Web 2.0 and Web 3.0 is accountability. With the former, you have no idea what your data is used for; you can be tracked, spied on, etc. through your data. You have no privacy with web 2.0, a typical wall with ears and eyes. There are no checks and balances. The challenges with web 2.0 (data ownership, privacy issues, accountability) can be resolved with web 3.0.
The Linked list — the meeting point between decentralized data and artificial intelligence
The linked list is a data structure (12 lines of code in python). A linked list is a simple object that is linked to the next one just as the name implies. How does the link list help move to web 3.0? Imagine storing single information or data across a million computers or nodes. This information or data represents a link list. Or imagine storing a transaction that occurred between two people on these nodes and store it in the way that no one can modify these transactions. To modify this information, we need many people on different supercomputers to solve complex computational mathematical problems. Now, for anyone to modify anything on any of these nodes or supercomputers, the person must have more than 50 percent control of all the nodes spread across the globe.
The term for this is 51% attack and this is almost impossible. For this attack to be possible, you will need more computing power than all the computing power on the supercomputers combined. What this means that no one can change the data on the linked list. This linked list that no one can have full control of is called the Blockchain. On the blockchain, these links and nodes are called blocks and the chains are pointing to nodes with transactions in it.
The decentralized data structure (Blockchain), artificial intelligence and Cartesi
The blockchain is a data structure that is immutable and decentralized. There is a trustless system existing with centralized systems whereby you cannot trust them completely with your funds and your assets. The advent of the data structure called the blockchain brought in a system whereby you don’t need a third party to verify any transaction.
How can we store large data on the blockchain? This can be achieved through the Intra planetary file system (IPFS) which is a network of computers. This will replicate the data across several nodes. One can also incentivize these nodes. Just like a neuron in the brain, it is an artificial neuron structure that mirrors the human neuron in deep learning. Now, this artificial neuron can learn any kind of function (linear and nonlinear function).
With deep learning revolution, the artificial neuron structure does everything that humans are good at (image recognition, driving cars, etc.) and sometimes even better. The artificial intelligence model will combine IPFS, Ethereum, and deep learning to decentralize and automate data. IPFS will store the data, Ethereum will build applications for smart contracts so that the data will not be stored on a central server and you can use deep learning to learn from that data.
Cartesi will power a decentralized internet where gamers can truly own their games and provide marketplaces for decentralized data scientists. Decentralized data scientists will be able to provide verifiable computation and businesses can outsource optimization tasks to experts. With Cartesi, these tasks can be incentivized as long as they are proven to be correct.