As artificial intelligence (AI) quickly works its complex magic in one sector of the economy after another, the need for computing resources to support all this machine intelligence is becoming more and more urgent.
Training a model like ChatGPT costs over $5 million, and running an early demo of ChatGPT, even before usage has increased to its current level, costs OpenAI about $100,000 a day. And AI is more than just text generation; the application of AI to solve practical problems in various industries requires similar large neural models trained on various types of data – medical, financial, customer information, geospatial, etc. Moving beyond the limitations of the current artificial intelligence of neural networks towards systems with a higher level artificial general intelligence will almost certainly require even more intensive computing.
It is only natural that a small but growing number of crypto miners are now looking for ways to use their own computing infrastructures to help drive the AI revolution.
Bitcoin Maximalists, From Bernie Madoff to Bankman Fried, Checked Out
Bitcoin (BTC) mining remains a profitable business. Mining other cryptocurrencies can also be profitable, but this is a rapidly changing landscape. Ethereum (ETH) miners, for example, were hit hard at the end of last year when the Ethereum network moved from proof of work to proof of stake.
The economic and technical situation in the crypto space over the past two years has prompted a growing number of cryptocurrency mining organizations to explore the potential of using their mining power for other purposes such as high performance computing and, in particular, artificial intelligence.
The specific computing hardware required for high-performance computing (HPC) or AI processing is often different from what is optimal for cryptocurrency mining. But buying servers is usually not the hardest part of setting up a mining farm. Providing power, cooling, security, and other physical infrastructure requires a lot of expense and effort, and it all stays pretty much the same whether you’re hosting lightweight GPUs suitable for ETH mining, or high-capacity GPUs. RAM suitable for training AI models. .
Mining company Hut 8 has taken the lead by leveraging its formerly mining computing power for machine learning and other HPC applications. Hive Blockchain has been doing the same for some time now, populating its servers with processor cards that “can be used for cloud computing and AI applications, as well as rendering for engineering applications, in addition to scientific simulation of fluid dynamics.”
Perhaps most interesting is the ability for miners to redirect their computing resources to AI in a way that stays entirely within the blockchain space—using them to run AI processes hosted on blockchain-based decentralized networks. A number of AI projects related to native altcoins, such as Fetch.ai (FET), Ocean (OCEAN), Matrix AI Network (MAN), Cortex (CTXC) and my own SingularityNET project (AGIX), and its various ecosystems provide this opportunity. projects such as NuNet (NTX) and the new registry-insensitive HyperCycle blockchain. AI-related altcoins performed well in the first half of 2023 as the market realized the potential of decentralized AI software.
Should Bored Ape buyers be legally entitled to a refund?
Even before the release of the Bitcoin White Paper, it was clear that the confluence of distributed computing, strong encryption, and decentralized control had wide applications that went beyond the financial. That’s why we have blockchain projects in areas that cover almost all vertical markets – medicine, supply chains, games, robotics, and so on. As AI dominates each of these business areas, the decentralization of the software and hardware behind AI will become a critical aspect of the decentralization of the global economy. The repurposing of a piece of cryptocurrency mining hardware to run AI processing, some of which is included in AI-focused cryptonetworks, will increasingly become part of history.
If a non-trivial piece of global AI processing ends up being performed on crypto-mining facilities, this could have implications beyond the financial ones. Cryptocurrency mining rigs are based in different legal jurisdictions and are owned by many different parties. A globally distributed AI network distributed across cryptocurrency mining rigs will be much more difficult to centrally control governments or other parties than an AI network centered on server farms owned by Big Tech (the current AI default). Whether this is good or bad in terms of AI ethics depends on your assessment of the nature of big tech and big government.
This article is for general informational purposes and is not intended and should not be taken as legal or investment advice. The views, thoughts and opinions expressed here are those of the author only and do not necessarily reflect or represent the views and opinions of Cryptooshala.
Credit : cointelegraph.com
