Crypto mining is still a profitable endeavor, but these days it has become more competitive than ever before. That's because mining demands more intensive computing power, as the biggest players continually invest in powerful hardware to ensure they can solve those "proof-of-work" calculations more efficiently than anyone else.
It's worth pointing out that mining isn't as simple as buying a few ASIC rigs, switching them on, and walking away, either. Crypto mining involves a very delicate balancing act with multiple moving parts. In addition to the ongoing operational costs (such as sky-high electricity bills), there are unexpected downtime and repairs that can impact profitability, along with product shipment delays that were considerably exacerbated during the COVID pandemic, when computer chips became an extremely valuable and hard-to-find commodity. There are also ever-changing regulations that vary from country to country, which can make life difficult for those investing in mining hardware.
Then there's the volatility of the crypto markets. From early 2022 to mid-2023, the crypto industry experienced one of its harshest winters on record. The price of Bitcoin suffered a precipitous drop from its all-time high of $69,000 to as low as $17,600 at one point, eating away at the profitability of mining operations. And now, even as crypto's bullish sentiment returns and Bitcoin hits a new all-time high, further changes are afoot. With the imminent arrival of Bitcoin's next "halvening," where mining rewards will be slashed in half, there is speculation that this event will have a significant impact on the mining industry, and many believe that smaller mining firms and individuals will be put out of business.
The crypto mining industry is looking less and less attractive, not only for those who are considering getting involved but also for those who are already participating. Luckily, the sudden emergence of artificial intelligence as a force to be reckoned with has created enormous demand for a source of computing and, with it, a new opportunity for crypto miners and anyone else who wants to get involved.
Crypto Miners Pivot to AI
The switch from crypto mining to AI computing kicked off in earnest in 2022 when an alternative blockchain network, Ethereum, announced plans to switch from its PoW consensus algorithm to a more energy-efficient "proof-of-stake" mechanism that does away with the need for electricity-hungry servers and computer chips.
As a result, the enormous Ethereum mining industry that had built up around the network was put out of work almost overnight. Given the ultra-competitive nature of the Bitcoin mining industry, many Ethereum miners dismissed switching to that cryptocurrency and instead began looking for alternative markets where they could put their powerful computing hardware to use.
Many miners use specialist ASIC chips for their number-crunching operations, and unfortunately, these aren't very efficient at other computing tasks. However, Ethereum miners could still use the rest of the infrastructure they had built, including their GPUs, cooling systems, security, and access to cheap sources of energy, to move into new markets, and few are as promising as AI.
With the sudden emergence of OpenAI's ChatGPT and generative AI in late 2022, the demand for AI computing has increased dramatically as almost every business races to take advantage of productivity-enhancing technology. Some mining-focused companies have already successfully pivoted, with one of the most notable being the Canadian company Hive.
Hive entered the crypto mining sector in 2017 and was initially very successful, building up a comprehensive infrastructure based on GPUs, which, fortunately, can be repurposed for different computing tasks. Hive's executive chairman, Frank Holmes, told The Guardian that the transition wasn't easy, as the company had to invest a significant amount of money in buying new servers to host its GPUs, as well as upgrade its networking infrastructure. Nevertheless, it managed the pivot so successfully that in June last year, it even saw fit to change its name from "Hive Blockchain" to "Hive Digital Technologies" to reflect its new emphasis on AI.
The Evolving AI Industry
What's most compelling about the AI industry is that it's evolving rapidly, and while the biggest advances have so far been made by centralized companies like OpenAI, with its proprietary GPT Large Language Models, there has been a concerted effort by the technology industry to build open-source alternatives. For instance, Facebook's parent company, Meta Platforms, has developed the open-source Llama 2 model as an alternative to OpenAI's GPT 4, and it has already been adopted by hundreds of enterprises to power various applications.
Moreover, the recent events at OpenAI, where its CEO Sam Altman was unceremoniously fired before being rehired just days later, raised concerns about the reliance on centralized AI development. Srinivasan Balaji, the former chief technology officer at Coinbase, has emerged as a big advocate for decentralized AI systems. Moreover, the requirements for compute infrastructure are so intense that Nvidia, the world's top GPU maker, has admitted that it is unable to keep up with customer demand. There are problems around fragmentation in the industry, too, with the vast majority of compute resources provided either by cloud infrastructure providers such as AWS, Microsoft Azure, and Google Cloud or else housed in on-premises systems. This means that AI systems generally cannot communicate with one another, limiting their potential.
If the AI industry is to live up to its potential and solve questions around demand and accessibility, then it needs to embrace a more decentralized approach, where anyone can access the compute resources needed and where AI systems can easily talk to each other. A decentralized approach will help to accelerate progress in AI and improve their intelligence, creating a new era of "artificial general intelligence," or AI systems that can perform multiple tasks simultaneously, instead of systems like ChatGPT, which only specializes in a single task, content generation.
Decentralized AI
The emergence of AGI has opened the door for projects like HyperCycle, which is building a ledgerless blockchain architecture that aims to democratize AI with a fast and secure network for AI computing that anyone can access.
HyperCycle has created a Layer 0 infrastructure platform that powers extremely rapid, low-cost transactions between diverse AI agents spread across a distributed but interconnected network in which anyone can participate. The idea is that these distributed nodes can work with each other to power AI computing tasks collectively. In this way, HyperCycle is building an Internet of AIs that will provide access to powerful yet low-cost compute resources as an alternative to today's siloed AI industry.
Some key innovations have helped to make HyperCycle's goal achievable. For instance, it is collaborating with a company called Penguin Digital to build a service called HyperPG to connect its network beneficiaries and enable them to make use of the abundant, low-cost hydropower resources in Paraguay so it can provide an environmentally-friendly energy source for AI computing. In addition, the company has created its very own hardware, the HyperAiBox, which is a plug-and-play device that anyone can purchase to set up a node and start contributing resources to the network, earning rewards for doing so. Of course, it's still possible to participate in its network with other types of hardware, such as GPUs, meaning its nascent AI network is a viable alternative for anyone looking to switch from crypto mining.
A Much Bigger Opportunity?
Decentralized AI is not just a technological shift but also a philosophical one that challenges the status quo of AI development. Until now, the AI industry has been dominated by a handful of corporations with the resources to build up massive data centers that can provide the vast power needed to train and run powerful generative AI models.
On the other hand, decentralized AI promotes an alternative—a shared, collaborative network where resources are pooled and made accessible to anyone who needs them. Such networks are open to anyone who wants to participate, with the only requirements being an internet connection and some kind of hardware that can contribute to compute resources. It's still early days in the rise of decentralized AI, but the approach has a lot of promise, with its potential to level the playing field and give smaller companies and developers an opportunity to participate in AI development and benefit from its advances.
The rise of the decentralized AI industry has many similarities to the early days of crypto mining, providing opportunities for participants to cater to a rapidly growing demand for compute resources. However, it brings new advantages, with entities such as HyperCycle building networks that are far less environmentally damaging. It's also much more useful, serving many different industries, as opposed to mining, which only benefits one. These advantages mean that decentralized AI may ultimately have much greater staying power than the crypto mining industry, meaning that it's an opportunity that can potentially persist for many years to come.
