Profile Photo

AI

  • Public Group
  • 2 months, 1 week ago
  • 5

    Posts

  • 2

    Members

Machine Learning’s Role in Sustainability for Digital Assets

Home Forum AI Machine Learning’s Role in Sustainability for Digital Assets

  • Author
    Posts
  • Tamil
    Tamil

    May 2021 has been a gift for every viewpoint in the Digital Assets market.

    The “its a Ponzi scheme” side of the debate will point to wild swings in valuation on an incredible 10,000 + crypto assets (as listed on coinmarketcap.com). The “its going to change the world” supporters point to the huge recovery late in May as large funds and “whales” buy as if Bitcoin was on sale.

    May 2021 has indeed been volatile, but for a technology heading into its twelfth year, this is standard operating procedure for old hands in the industry – even the old hands who have only just left High School.

    However, one latent theme that has bubbled up and moved markets (to the tune of 0.5 Trillion USD in a matter of weeks) is the subject of Bitcoin mining power usage.

    Enter Elon Musk

    Elon Musk is undoubtedly one of the most influential figures in Digital Assets. His support for Bitcoin was recently undercut by his tweets about the effect Bitcoin mining is having on the network’s carbon output. Along with comments from China, this forced a large market correction.

    After meetings with US based mining groups, and an intervention by Michael Saylor of MicroStrategy, sustainability in Digital Assets is now de rigueur and driving markets higher as we head into June 2021.
    So, what is the actual situation regarding sustainability in Digital Assets and how can Machine Learning transform our understanding and ability to ensure Digital Assets do not destroy the planet.

    Carbon Content in Digital Assets

    Proof of Work is a protocol enshrined in the pseudononymous paper from Satoshi Nakamoto, whereby a permissionless network of computers solve huge mathematical problems to secure the network and create a trustless, peer to peer payment network.

    Whilst brilliant in its design, its downside is the arms race underway to mine the network. The fact that the Bitcoin protocol does not distinguish between the carbon content of the electrons used to power the network, means that coal powered grids in China can participate alongside emerging “green” miners in North America and Europe with no penalties.

    The main strategy for other projects is to use different protocols to secure the network, such as Proof of Stake, Proof of Authority and so on. These techniques do not require massive amounts of computation (power) and have other exciting benefits such as speed of computation (measured in Blocktime).

    A very exciting project called Internet Computer (ICP) is 1,500 times faster than Bitcoin and aims to increase Block Time another 400 times faster yet (being able to run at 1,000 blocks per second – by comparison, Bitcoin produces one block every 10 minutes).
    However, Bitcoin has a massive advantage in terms of being the gold standard for Digital Assets and much like standards wars in the past (VHS versus Betamax), Bitcoin is unlikely to relinquish its spot as the preeminent Digital Asset anytime soon, despite its opaque power infrastructure.

    Machine Learning and Bitcoin Power Use Cases

    As with any domain, there are many use cases that can leverage the awesome power of Machine Learning. But few domains are as data rich as Digital Assets. Public BlockChains are open and anyone can access the incredible wealth of immutable data. Bitcoin is a data scientists dream.

    As a subset of (valuable) research opens up into the sustainability of the network, yet unknown facts about the network will lead to new and exciting research for anyone with an enquiring mind, a Bitcoin node and a decent wireless connection including:

    Bitcoin price effects on mining hashrate;
    Network latency;
    Geographic distribution of miners;
    Economic incentives of mining pools;
    Future network difficulty projections;
    Efficiency of new mining rigs;
    Etc

    Machine Learning will enable new analytics to emerge that:

    detect new signals in the noise of data in the eco system;
    Enable new algorithms to make mining more predictive;
    Empower miners to be smarter in the way power is deployed and priced.
    Machine Learning analytics will inexorably lead to new operational techniques that will drive down power usage, create arbitrage opportunities for global power markets and lead to a more enlightened approach that does less harm to the planet.

    Machine Learning based insights will drive down power usage, without losing the many attractive characteristics of the world’s first fully decentralised, peer to peer payment system.

    As the application of these new technologies becomes as important as their development, the value of this data will rise. Anyone able to unlock the mysteries and devise new strategies in Bitcoin mining will create huge value, and help reduce the power consumption of the network faster than any ASIC chip designer.

    Working smarter, not just harder, is never a losing strategy. The Machine Learning age in Bitcoin is just beginning.

You must be logged in to reply to this topic.

Up
Top Ranked Users

[joinup_core_top_members columns=”1″ space=”no” max_members=”3″ behavior=”columns” columns_responsive=”predefined”]