IOTA is a unique project in the crypto industry that has set out to lay the foundations for new emerging markets in not just reinventing the distributed ledger, but building the infrastructure for the future landscape of the Internet of Things. The tangle, IOTA’s data coordination model, is a directed acyclic graph adapted to function as a real-time data transport layer and decentralized information storage. Neural networks as such approximate functions on the weight matrices of directed acyclic graphs and IOTA’s tangle, likewise, deals with probabilities in gradually aligning itself with factual reality in a training-based model.
Neural networks are structured as directed acyclic graphs.
In tackling the technical challenges of IoT, IOTA transcends the limitations of the common blockchain paradigm, substituting with a lightweight and flexible architecture where activity on the network itself is what secures the network, while participants are essentially being rewarded with zero fees (whether issuing transactions, transmitting data, or running applications). The IOTA network is essentially collaborative in a way that anybody who wants to use the system empowers the network in the process (rather than any given group of actors on the network), and is incentivized more by cost-saving efficiency and access to validated, high-quality data than immediate profits.
IOTA’s recently announced and eagerly anticipated Qubic (with details expected to be released on the 3rd of June) is an extension of the core protocol (similar to what TCP/IP is to the internet protocol suite) that enables smart contract functionality and outsourcing of computations. It appears to fuse with the data marketplace, which provides fine-granular data streams from devices and sensor nodes from around the globe, such as environmental, agricultural, and anonymous healthcare data, thereby facilitating the shaping of an open, permissionless market for data-driven insights.
Blockchains, in comparison, function in a strictly deterministic fashion and are subject to fluctuating fees, while blockchain-bound smart contracts are blind, deaf, and dumb in that they are static objects that are triggered only when they are being fed the data necessary (think vending machines or trap doors). Furthermore, on-chain computations are prohibitively expensive (making estimative calculations practically impossible). IOTA on the other hand conditions an entirely different environment and playground of yet unexplored possibilities.
The Bayesian methods of statistical inference IOTA employs are usually encountered in the domains of machine learning and AI research (Markov Chain Monte Carlo, weighted random walks, etc.) and the network not only becomes stronger and faster in its capacity to scale exponentially as it gets populated with activity, but is also capable of differentiation, orientation, dynamic organization, and wise decision-making at scale.
Qubic is expected to empower users and developers to functionally define smart contracts (similar to how Cardano uses Haskell to express smart contracts), adjusting their own parameters of reasonable expectations and degrees of accepted risk, which would traverse all the data points and resources available on the tangle, unencumbered by block times and network fees (aside from what owners of said data might be charging to open their streams up, but data may be fetched from certain sources at no cost), while simultaneously acting as their own oracles.
The efforts that have gone into conceptualizing and developing Qubic stretch as far back as 2012 and despite having been only recently announced to the public officially, there have been clues about it in the past:
“Our approach to Oracles will make it possible to directly form oracles on top of IOTA. This is unlike anything that’s currently being used in this space. With our Oracle platform we not only aim at connecting the physical world with IOTA; but we also want to enable true interoperability with all other Blockchain platforms.”
The DIKW pyramid, used to illustrate the interdependent relationships between raw, unstructured data, the information data can provide when contextualized, the working knowledge that consequently arises from information, and eventually, the applied knowledge of wise decision making this enables.
Timestamps on the tangle will be determined in terms of confidence intervals of reasonable accuracy, since there are only flows and different chains of interdependencies between events and transactions on the tangle and not a universal ordering of events as on a blockchain.
Outsourced computations will enable the running of applications in a massively parallel, resource-sharing system, which opens the horizons for intriguing future possibilities, as recently published studies have demonstrated the capacity of machine learning techniques in predicting chaotic behavior.
As Quanta magazine writes, “Parallelization allows the reservoir computing approach to handle chaotic systems of almost any size, as long as proportionate computer resources are dedicated to the task.”
David Sonstebo comments on IOTA Tangle, “The goal is indeed to get mining pools to switch over to providing a useful service (machine learning) while paid in iotas. A lot of mining farms are struggling these days, so this is a very win-win situation.”
Instead of brute forcing random numbers in a technological arms race, miners could be put to do useful work. However, it must be understood that the pragmatic nature of the IOTA enterprise is such that it implies a step-wise process of gradual optimization and not a prescriptive agenda to be rolled out and hardcoded on a blockchain. As such, the usual criteria habitually applied to the fabric of blockchain platforms are often unrelated.
IOTA continues to set the standard for innovation in off-chain crypto developments and Qubic is its latest venture into data-driven smart contracts. In a marketplace full of speculation, IOTA has become an industry leader to be reckoned with.