...
+7 (495) 415-84-68 info@gbig.ru.com Moscow, Presnenskaya Embankment, 12
mining companies

The Great Transformation: How Mining Companies Outmanoeuvred Bitcoin

October 29, 2025 | Reading time: 9 minutes | GBIG Holdings

Prologue: When Mathematics Stopped Converging

April 2024 was supposed to be another milestone in the history of bitcoin – the fourth halving, an event that once every four years divides the reward of miners in half. From 6.25 BTC to 3.125 BTC per block. The classical model predicted further development: falling incomes, the release of weak players, the consolidation of the industry, and then a gradual rise in the price of bitcoin, which will compensate for the losses.

But something didn't go according to the script.

By September 2024, Bitcoin mining profitability had fallen by more than 7 percent. Hashprice — miners' revenue per unit of computing power — dropped from a comfortable 60 dollars per petahash per day to a painful 52–53 dollars. Operating costs per bitcoin reached 82,162 dollars; including non-cash expenses such as depreciation brought the figure to 137,018 dollars.

With these numbers, the industry had to drown: small players to capitulate, big players to grit their teeth and wait for better times.

Instead, something surprising happened: mining company shares began to deliver returns that did not merely exceed Bitcoin's rise but eclipsed it. While in 2024 BTC rose by a modest 10 per cent, shares IREN soars 300 percent in a year and 400 percent in 12 months.. Core Scientific, which in late 2022 filed for Chapter 11 bankruptcy, has become an attractive acquisition target worth 9 billion dollars for the AI hyperscaler CoreWeave.

So what happened? Where did that money come from when Bitcoin mining itself was bringing less and less money?

The answer turned out to be simple and complex at the same time: miners are no longer just miners.

Anatomy of crisis

The Perfect Storm

To understand the scale of the transformation, one must first understand the depth of the crisis in which the industry finds itself in 2024.

Strike one: halving. On April 20, at block height of 840,000, the block reward halved. This is not news – halvings occur regularly, they are written in the very code of bitcoin. But each time, the industry hopes that the price rise will compensate for the loss. This time, the compensation did not happen quickly enough.

Key figures: Before halving, from March to the end of April, miners' returns were in the range of 0.10 to 0.12 dollars per terahesh per second, and then collapsed to 0.04 to 0.06. Simple arithmetic: minus 50-60 percent of revenue overnight.

The second blow: rising difficulty. The computational power of the bitcoin network, the hashrate, not only did not fall after halving, it continued to grow. In December 2024, the network reached an all-time high of 801 exahes per second. By April 2025, it was already 924 EH/s.

This means that miners kept plugging in new hardware despite falling profitability. Why? Because stopping was even more dangerous, because if you fall behind technologically, there's no going back.

Strike three: energy costs. Electricity has always been a major expense for miners. But in 2024, amid geopolitical instability and rising energy prices, this factor became critical. Even in regions with relatively cheap energy, such as Russia, where the average cost of mining a single BTC was projected at $24,000, business margins have become fragile.

Strike four: debt load. According to The Miner Mag, the total liabilities of 25 public mining companies rose to $4.6 billion in the fourth quarter of 2024, then to $1.5 billion in the second quarter of 2025. This is not an expansion in the comfort zone - it is a forced attraction of capital in order not to drown.

Moment of truth

In the summer of 2024, according to analysts CoinSharesIf current conditions are maintained, direct investment in bitcoin will be four times more profitable than investment in mining in the next four years.

Four times. Imagine you're building a business, investing in equipment, infrastructure, energy, and in the end it would be easier to just buy bitcoin and hold it.

For public companies with shareholders, it was an existential crisis. Marathon Digital, CleanSpark, Riot Platforms, Core Scientific, Hive Digital, IREN – all these giants of the industry faced one question: what next?

And they found the answer where no one expected it.

The Great U-turn

AI: The New Gold Rush

At the same time that the mining industry was suffocating, another industry was experiencing a boom the world hasn’t seen since the Internet bubble era of the late 90s.

After the launch of ChatGPT in November 2022, demand for computing power for neural network training and operation exploded. Companies like OpenAI, Anthropic, Google, Meta rushed to buy GPUs — Nvidia GPUs, the only chips that can cope with the giant matrix computing required for modern language models.

And there was an epiphany.

Mining companies had three things that the AI industry desperately needed:

  • Infrastructure: data centers with powerful cooling, sustainable power supply, high-speed Internet.
  • Energy contracts: Long-term agreements to supply cheap electricity, often from renewable sources.
  • Expertise: Experience managing high-performance computing 24/7.

Of course, you can’t just take an ASIC miner for bitcoin and force it to train neural networks. Application-Specific Integrated Circuit is a “specialized integrated circuit” designed exclusively for the calculation of SHA-256 hashes. For AI, you need graphics processors, preferably top models like the Nvidia H100, H200 or the latest Blackwell B200.

But here's what's important: the physical infrastructure -- buildings, cooling systems, electrical substations, networking equipment -- can be repurposed. Needham, capital expenditure on high-performance computing data centres is 8–10 million dollars per megawatt, excluding GPUs, compared with just 300–800 thousand dollars per megawatt for mining farms, excluding ASICs.

It's 10-30 times the difference.

Mathematics of the New World

Let's compare returns:

Bitcoin Mining (2024-2025):

  • Income: $0.04-0.06 per TH/s per day after halving
  • hashprice: $45-60 per PH/s per day (depends heavily on BTC price)
  • Operating expenses for 1 BTC: $82,162
  • Income Volatility: Extremely High

AI/HPC computing:

  • Revenue: $8 to $10 million a year in contracts for a few hundred GPUs
  • Yield per kilowatt hour: up to 25 times higher than crypto mining
  • Contract Model: Fixed Long-Term Agreements (5-12 years)
  • Income Volatility: Low (independent of the crypto market)

A simple example: IREN deployed just 248 GPUs in early 2024 and signed the first contract to develop AI. By mid-2025, the number of GPUs had grown to 4,300, and AI cloud services were bringing in $3.6 million quarterly. At the same time, the company continued to mine bitcoin, mining 1,514 BTC in the third quarter of fiscal 2025.

Hive Digital Technologies has gone even further, with revenue from AI and HPC hosting tripling to $10.1 million in fiscal year 2025 – nearly 9 percent of the company’s total revenue. The target for 2026 is 100 million from AI.

But the real revelation was the story of Core Scientific.

Revival from the Ashes: The Story of Core Scientific

End of 2022. Core Scientific is filing for Chapter 11 bankruptcy. The reasons are standard: the fall in the price of bitcoin during the crypto winter, the huge debt load, the inability to service loans. The company owed about a billion dollars.

Lenders began seizing the mining equipment used as collateral, a classic miner capitulation scenario of writing off losses and selling off assets piecemeal.

But Core Scientific took a different path.

January 2023

Bankruptcy Court approves restructuring plan

Beginning 2024

The company returned to the Nasdaq exchange

June 2024

Signed a 12-year deal with CoreWeave worth $3.5 billion

October 2024

The contract has been extended by another 120 megawatts (total capacity of about 500 MW).

July 2025

CoreWeave acquires Core Scientific for $9 billion

Nine billion dollars for a company that went bankrupt two and a half years ago.

A synergy that didn't exist

“AI Potential to Drive Exponential Growth”

Adam Sullivan, CEO of Core Scientific, September 2024

What exactly creates this synergy? Why are the mining companies in the right place at the right time?

Factor 1: Flexibility of energy consumption. Russell Cann, the development director of Core Scientific, drew attention to the critical difference: Bitcoin mining can be turned off at any time without catastrophic consequences. No AI computation. Training a neural network that is suddenly interrupted can nullify weeks of work.

This means that miners already know how to balance electricity consumption by participating in power grid balancing programs. Riot Platforms, for example, earned $12.4 million in the third quarter of 2024 from equipment outages alone during periods of peak demand (plus $19 million in the first half).

Factor 2: Remote location. Many mining farms are located in places with excess electricity generation – hydroelectric power plants in Canada and Norway, geothermal sources in Iceland, associated petroleum gas in Texas. These are ideal locations for energy-intensive AI data centers, which do not need proximity to users, but cheap electricity is critical.

Factor 3: Environmental profile. Many miners have long switched to renewable energy – not out of altruism, but out of pragmatism: it’s cheaper. Hive Digital Technologies, for example, runs exclusively on renewable energy in Canada, Sweden and Paraguay. For AI companies that are increasingly concerned about the carbon footprint, this is an important advantage.

Portraits of Transformation

IREN: 248 GPUs to 500 million in revenue

The story of Australia's IREN (formerly Iris Energy) is the story of how a small bet turned into a strategic pivot.

Beginning 2024

Company buys 248 Nvidia GPUs.modest pilot program

February 2024

The first contract to develop AI based on these GPUs

Mid-2025

There are 4,300 GPUs installed and running

August 2024

IREN Announces Massive $193 Million Investment in Nvidia Blackwell B200 GPUs, Then Increases Rate to $674 Million, Acquiring 12,400 New GPUs

Forecast: Expected annual revenue from AI cloud services – $500 million by early 2026

Current indicators in the third quarter of fiscal 2025:

  • Bitcoin Mining: 1,514 BTC (up from 1,232 BTC in the previous quarter)
  • AI revenue: $3.6 million quarterly, or $14.4 million year-on-year
  • Stock growth: plus 300 percent for 2025, plus 400 percent for 12 months

The company is building two flagship facilities:

  • Horizon 1 Texas: $300-350 million, 200 kW per rack, direct chip cooling for Nvidia Blackwell
  • Sweetwater: 2 GW of power, able to accommodate more than 700,000 Blackwell liquid-cooled GPUs

The dark side: In October 2024, a class-action lawsuit was filed against IREN alleging that the company misled investors about the readiness of the Texas facility. The allegations have not yet been proven, but they show how aggressively the companies are moving in that direction.

Hive Digital Technologies: 100 million by 2026

Canada's Hive has taken a more cautious but consistent path. The company began rebranding back in mid-2023, changing its name from Hive Blockchain to Hive Digital Technologies, signaling an expansion of focus beyond cryptocurrencies.

Results for fiscal 2025:

  • Revenue from AI and HPC Hosting: $10.1 million (three times year-on-year)
  • Share of AI in total revenue: about 9 percent
  • Bitcoin mining: hashrate rose to 10.4 EH/s by May 2025 (up 58 percent month-to-month)
  • Plan to reach 25 EH/s by the end of 2025

But the main thing is an ambitious goal: $100 million in annual AI revenue by 2026.

How do they plan to do that? Investments in Nvidia GPU clusters, expansion in Paraguay (where electricity is virtually free thanks to hydropower), and the creation of a subsidiary of Buzz HPC to develop sovereign AI data centers for Canadian government agencies and enterprises.

Hut 8: Bitcoin treasury plus AI compute

Hut 8 has opted for a hybrid model where mining remains the core, but AI adds diversification.

In the first quarter of 2025, the company mined 167 BTC – a significant decrease from 716 BTC in the same period of 2024. The culprit is halving. But the Hut 8 isn't panicking. The company has a balance sheet of 10,273 BTC, making it the ninth largest corporate bitcoin holder in the world.

That's the storage position: mine bitcoin, don't sell it, wait for the price to go up, and in the meantime, the price doesn't go up as fast as you'd like, make money on AI.

Marathon Digital and Riot: Traditions

Interestingly, not all miners rushed into AI.

Marathon Digital (MARA) and Riot Platforms, the two largest public miners, remain largely committed to bitcoin.

MARA owns the largest bitcoin treasury among miners: 50,000 BTC on the balance sheet, second only to MicroStrategy among all public companies. In December 2024, Marathon achieved a target hashrate of 50 EH/s, improving the energy efficiency of miners to 20 J/TH.

Riot mined 1,530 BTC in the first quarter of 2025, earning $142.9 million from mining - twice as much as a year earlier.The company owns 19,225 BTC, fourth among corporate holders.

Their strategy is to double their bet on Bitcoin, believe in long-term price growth, and remain the most effective miners. While others are experimenting with AI, Marathon and Riot are buying up equipment, ramping up capacity, and squeezing out less efficient competitors.

Who is right, time will tell.

The Dark Side of the Boom

Debt bomb

The splendid numbers of billions in stock and contract growth hide an inconvenient truth: the entire U-turn is financed in debt.

Public mining companies raised more than $4.6 billion through debt instruments and convertible bonds at the end of 2024, with an acceleration in 2025.

Example of TeraWulf: The company issued convertible bonds with a coupon of 7.75 percent, meaning an annual interest expense of about $250 million. This is despite the fact that the total revenue of the company in 2024 was only 140 million.

IREN uses zero-coupon convertible bonds – a tricky scheme where interest is not paid immediately, but bonds are converted into shares at a certain price, blurring the share of existing shareholders.

Why do companies choose debt rather than issuing new shares directly?

  1. The market rewards growth: As stocks rise, convertible bonds seem like a cheap way to finance them.
  2. Faith in AI: Investors are ready to give money for AI projects at lower rates than for mining
  3. Deferring dilution: The conversion occurs later when (theoretically) the company will be worth even more.

But what if the music stops? In 2022, during a bear market where bitcoin fell 70 percent, lenders seized miners’ equipment used as collateral. This is how Core Scientific went bankrupt.

Technical risks

Repurposing mining infrastructure to AI is not just a switching switch; AI computing requires:

  • Higher power density per rack (200 plus kW vs 30-50 kW in bitcoin mining)
  • Liquid cooling (immersion or direct chip)
  • High-speed networks (InfiniBand, 400Gbit Ethernet)
  • Specialized software for orchestration of calculations

Adam Sullivan of Core Scientific admitted that not all 1.2 GW of infrastructure can be repurposed - the technical requirements are too different.

Competition

Miners aren't the only ones who want to make money off the AI boom. Amazon (AWS), Microsoft (Azure), Google (GCP) They are building their own data centers for AI, and by orders of magnitude larger than what former miners can offer.

The competitive advantage of miners—cheap energy and rapid deployment—may be temporary. Once traditional cloud providers build enough infrastructure in regions with cheap electricity, price pressures will increase.

Regulatory risks

Nor should we forget about regulation. In 2024, governments around the world have tightened their focus on the crypto industry. The U.S. planned to impose an additional tax of 30 percent specifically for miners (although they have so far refused). The European Union is introducing new environmental standards. Russia legalized mining, but with strict restrictions and mandatory reporting to Rosfinmonitoring.

The future of the hybrid model

Combining both approaches

The most interesting thing about the current reversal is that most miners are not abandoning Bitcoin completely; they are not closing ASIC farms or selling BTC reserves. Instead, they are building a hybrid model.

Some of the infrastructure continues to mine bitcoin, some of them are being converted to AI/HPC, and depending on the current returns, the power is being redistributed.

It's smart for a couple of reasons:

Hedging volatility. Bitcoin mining is a highly volatile business tied to the price of BTC. AI computing is a more stable, contract-based business. The combination gives a predictable base income plus the potential for a high upside if Bitcoin flies into the stratosphere.

Optimization of energy use. During periods of high electricity demand, miners can switch off ASICs and sell surplus energy or provide it to AI customers. When demand falls, they can restart mining.

Preserving optionality. No one knows what will happen to the AI boom in 5-10 years. Maybe it will continue exponentially. Maybe it will run into fundamental limits (already now the U.S. Department of Energy warns that data centers will consume 6.7-12 percent of the country’s total electricity by 2028 – this is a physical problem).

Does this threaten the Bitcoin network?

Critics rightly ask: if miners massively leave for AI, will the security of the bitcoin network suffer?

“Digital gold mining remains an easy way to quickly monetize excess electricity in remote or emerging markets, effectively subsidizing the development of data centers that were originally designed to be converted to AI and HPC needs.”

Analysts VanEckNathan Frankowitz and Matthew Siegel

In other words, mining becomes a flexible ballast that fills the gaps in the loading of the energy infrastructure. When the demand for AI computing is high, AI works. When demand falls, mining starts.

The hashrate of the bitcoin network, by the way, continues to grow despite all the talk about going into AI, which means that diversification does not mean rejection – it means addition.

Lessons for investors

How to Read a New Reality

For those considering investing in mining companies, the world has changed dramatically.

The old paradigm (until 2024):

  • Shares of miners = leverage on bitcoin
  • If BTC Rises 10 Percent, Stocks Rise 20-30 Percent
  • If BTC falls 10 percent, stocks fall 30 to 50 percent.
  • Key metrics: hashrate, cost of electricity, equipment efficiency

A new paradigm (2024 plus):

  • Miners’ shares = a mixture of crypto exposure plus AI infrastructure game
  • Growth does not necessarily correlate with Bitcoin
  • Key metrics: hashrate plus GPU number plus long-term AI contracts plus debt load plus energy agreements

What to look at

1. Share of AI revenue. The higher the percentage of revenue from AI computing, the more stable the business, but less exposure to Bitcoin.

  • IREN: Aims for AI Dominance (500 million out of approximately 520 million total revenue by 2026)
  • Hive: 50/50 target by 2026 (100 million AI, approximately 100-150 million mining)
  • MARA/Riot: almost 100 percent mining, minimal diversification

2. Quality of AI contracts. Not all contracts are the same. Long-term (5-12 years) agreements with large clients (CoreWeave, Google, Microsoft) are gold. Short-term or spot contracts are much riskier.

3. The cost of energy. This is a fundamental competitive advantage: Companies with energy access of less than $0.04 per kWh (Hive in Paraguay, IREN in Texas with renewables) have a structural advantage.

4. Debt load and repayment schedule. Companies that have amassed huge debts to fund the AI transition are vulnerable if revenue does not grow as fast as planned.

  • Debt-to-EBITDA ratio
  • Repayment Schedule (when you need to return the principal amount)
  • Conversion conditions (at what price bonds are converted into shares)

5. Execution risk. The promise of 500 million in revenue from AI is one thing: building data centers, installing GPUs, finding customers, launching on time — quite another. The class lawsuit against IREN is a reminder that not everything is going according to plan.

Currency is the new puzzle

How do you value mining companies now? It used to be simple: take the company's hashrate, multiply by the current hashprice, subtract operating expenses, add the value of BTC on the balance sheet. You get a rough estimate.

Now you need to evaluate as a tech infrastructure company:

  • DCF Model Based on AI Contracts
  • Multipliers to revenue (compared to cloud providers)
  • The Cost of BTC Reserves as an Option for Growth
  • Premium or discount for execution risk

Important: In 2021, many companies added "blockchain" or "crypto" to the name, and their stock soared with no real business change. Now a similar story with "AI" — added a couple of slides in the presentation about the plans in AI, and the assessment grew. You need to look at real contracts, real GPUs, real revenue.

What Does This Mean for Bitcoin

There is a certain irony in the fact that mining companies that literally create Bitcoin earn more when they switch to something else.

But if you look deeper, it's not a contradiction, it's an evolution.

Bitcoin does not need miners who depend solely on it. The protocol doesn’t care who mines the blocks – pure miners or hybrid AI/mining companies. It is only important that there is sufficient computing power to maintain the security of the network.

And as miners diversify, they become more resilient and less prone to boom-bust cycles in the cryptocurrency market, making the industry healthier in the long run.

What’s more, the infrastructure that’s being built for AI computing – huge data centers powered by cheap renewable energy – is an infrastructure that could return to mining at any moment if the economy turns around.

Imagine 2030. Bitcoin has gone through another halving, the reward is now 1.5625 BTC per block. But the price went up to $500,000. Suddenly, mining becomes very profitable again. Companies that have retained the hybrid model can instantly shift power back.

That is resilience.

Final answer

Why do mining companies start earning significantly more than Bitcoin’s returns offer?

Because they are no longer just mining companies.

They’ve become energy management companies, high-performance computing operators, cloud infrastructure providers for AI — with built-in options to mine digital gold when it’s profitable.

The Halving of 2024, which was supposed to be a crisis, turned into a catalyst for transformation. Those companies that realized this in time and acted decisively - Core Scientific, IREN, Hive - increased their value significantly.

Those who have stayed true to pure mining – Marathon, Riot – are betting on Bitcoin’s long-term growth and operational efficiency.

Both strategies have a right to life, but one thing is clear: the mining industry of 2025 is not the mining industry of 2020.


Frequently asked questions

Can small miners also switch to AI computing?

For small miners, the transition to AI computing is much more difficult than for large public companies. The main problem is capital expenditure. One Nvidia H100 GPU costs about 30-40 thousand dollars, and for a serious AI business you need hundreds or thousands of such chips. In addition, modernization of cooling systems, network infrastructure and the involvement of machine learning specialists are required.

However, there are alternative models: some small miners rent out their space and power capacity to larger AI companies, earning a steady income without having to buy expensive hardware on their own. This is an intermediate option between pure mining and a full-fledged AI business.

How is ASIC mining different from GPU computing for AI?

ASICs (Application-Specific Integrated Circuit) are specialized chips designed exclusively for mining cryptocurrencies with a specific algorithm, such as SHA-256 for bitcoin. They are incredibly effective at their narrow task, but absolutely useless for anything else.

GPUs (graphics processors) are universal chips capable of performing parallel calculations for a wide range of tasks: from rendering graphics in games to training neural networks. AI computing requires GPUs, as they handle the matrix operations required to work with large language models.

The key difference is that ASICs can't be reprogrammed for other tasks, and GPUs can, which is why mining companies can't just switch existing ASICs to AI — they have to buy new hardware.

Why Do Companies Prefer Debt Finance to Issuing New Stocks?

The choice of debt financing, especially convertible bonds, is due to several reasons. First, while a company’s shares are rising, convertible bonds can delay the erosion of existing shareholders’ shares. Investors will get shares later when the company is (theoretically) worth more.

Secondly, the market is now very positive about AI projects, so investors are willing to provide debt financing at lower interest rates than it would be for a pure mining business. The word “AI” in the presentation literally reduces the cost of capital.

However, this is a risky strategy. If revenue does not grow as fast as expected, or if the AI boom slows down, companies will find themselves with huge debts to service. The history of Core Scientific in 2022 is a vivid example of what happens when the debt burden becomes unbearable.

How long does it take to repurpose a mining farm for an AI data center?

The timing varies depending on the scale of the project and the existing infrastructure. For a relatively small object with a few hundred GPUs, the process can take 6-9 months. For large flagship data centers like Horizon 1 from IREN — 18-24 months.

The main stages include: engineering design and obtaining permits (2-4 months), modernization of power supply and cooling systems (4-8 months), installation of network equipment and racks (2-3 months), procurement and installation of GPUs (depending on the availability of chips, can take several months due to shortages), software configuration and testing (1-2 months).

The availability of Nvidia GPUs is often a critical factor – the demand is so high that companies are forced to place orders six months or a year in advance.

What environmental aspects are associated with the transition from mining to AI?

Interestingly, from an environmental point of view, AI computing is not necessarily greener than mining. Learning one large language model can consume as much energy as a few dozen households in a year. However, there are several important differences.

First, many mining companies already run on renewable energy sources (hydro, solar, wind), making them attractive partners for AI companies concerned about the carbon footprint. Second, AI data centers usually have higher energy utilization – they run at full capacity all the time, whereas mining can be less efficient due to fluctuations in returns.

On the other hand, rising electricity consumption in the AI industry creates serious challenges. The US Department of Energy forecasts that by 2028, data centres could consume up to 12 per cent of the country's electricity. This risks overloading power grids and could raise electricity prices for all consumers.

How is mining and AI computing regulated in different countries?

The regulatory environment varies widely across countries and continues to evolve. Mining is legal in the U.S., but some states have restrictions on power consumption. A planned federal tax of 30 percent on electricity for miners has not yet been enacted, but the idea periodically returns to political debate.

The European Union has strict environmental standards, especially after the adoption of MiCA (Markets in Crypto-Assets) regulation. Companies must disclose their carbon footprint and may face restrictions if they do not meet environmental requirements.

In Russia, mining was legalized in 2024, but with significant conditions: miners must register, comply with energy consumption limits and report to Rosfinmonitoring. The country ranks second in the world in terms of cryptocurrency mining after the United States.

China, once a leader in mining, imposed a blanket ban in 2021, leading to a mass exodus of miners to other jurisdictions. For AI computing, regulation is still less stringent, but there is a growing focus on data privacy and national security issues, especially in the context of geopolitical competition.

What are “sovereign clouds” and why are they important to miners?

Sovereign clouds are cloud computing infrastructures that are entirely located within a single country and are subject to its laws. This is critical for governments and companies working with sensitive data that cannot be transferred abroad for national security reasons or because of regulatory requirements.

For mining companies moving to AI, sovereign clouds are opening up a new market. For example, Hive’s subsidiary, Buzz HPC, specializes in building sovereign AI data centers for Canadian government agencies and businesses. Canada, like many other countries, wants critical computing to remain under national jurisdiction.

This is especially true amid geopolitical tensions and growing concerns about technological dependency. Governments do not want their AI infrastructure to be dependent on U.S. or Chinese cloud providers. Mining companies that already have the infrastructure on the ground and understand the local energy and regulatory environment are ideal for this niche.

Could the AI boom end the way previous tech bubbles ended?

The history of technology is full of examples of boom-bust cycles. The Internet bubble of the late 90s, the blockchain mania of 2017-2018, the current boom of artificial intelligence – all follow similar patterns: explosive growth of interest, rapid investment, inflated expectations, and then correction.

But there are important differences. AI, unlike many crypto projects in 2017, is already showing real utility. ChatGPT has hundreds of millions of users. Companies incorporate AI into real business processes. It’s not just a speculative asset; it’s a technology that’s changing the way we work.

However, the current estimates of many AI companies can be overstated. If revenue growth fails to meet expectations, or if technological constraints (e.g., physical limits on energy consumption) emerge, a significant correction is possible. For mining companies that have accumulated large debts for AI projects, this can be a serious problem.

The golden rule is that not all companies that claim to switch to AI are actually building sustainable businesses. Investors need to carefully analyze real contracts, real customers, and real revenue, rather than simply reacting to the mention of AI in press releases.


GBIG Holdings launches affordable mining project

The company GBIG Holdings announces the launch of an innovative project that allows individuals to participate in cryptocurrency mining through the acquisition of shares.

Key advantages:

Low entry threshold - from 5000 rubles

Professional management of GBIG experts

Full regulation of VTB Registrar, Central Bank of the Russian Federation and Federal Tax Service

Priority conditions for early investors

With the proceeds from the sale of shares, the company buys the best mining equipment and places it on sites with optimal tariffs, and the profits from mining are distributed to shareholders in the form of dividends.

Submit a preliminary application

At this stage, no investment is required, and the individuals included in the application book will benefit from the launch of the project.

Professional wealth management for high-net-worth investors

Entrust your assets to GBIG HOLDINGS experts with experience in traditional and digital markets. Tailored strategies for portfolios starting at 10 million rubles, with above-market returns.

Book a personal consultation

We provide complete confidentiality and a personalised approach.

Contact the experts at GBIG Holdings

Interested in investing in digital assets and innovative fintech projects? The GBIG Holdings team has expertise in digital asset management and can provide professional advice on building an investment portfolio.

Серафинит - АкселераторОптимизировано Серафинит - Акселератор
Включает высокую скорость сайта, чтобы быть привлекательным для людей и поисковых систем.