Crypto World
Exploring AI's Potential to Create Its Own Progeny
Introduction
Artificial intelligence (AI) is one of the most fascinating and disruptive technologies of our time. It is defined as a computer program capable of learning and improving independently, distinguishing it from other programs that follow fixed instructions. In this article, we will explore an intriguing aspect of AI: its ability to create other AI. Join us on this journey to better understand how AI can give birth to new artificial intelligence and what this advancement entails in the world of technology.
AI Creating Another AI
Since 2017, we have witnessed an intriguing phenomenon in the field of artificial intelligence: the ability of an AI to give rise to other AIs, often referred to as “AI offspring.” However, it is important to note that this process does not occur independently. In most cases, AIs that can create other AIs are specifically designed for this purpose and receive the necessary training to carry out this task. The current process raises the fundamental question of whether AI is capable of self-generation or if it will always depend on human direction for its evolution.
Supervised and Unsupervised Learning
To better understand this phenomenon, it is essential to delve into two approaches to learning in artificial intelligence: supervised learning and unsupervised learning. The former involves a process in which AI learns from a predefined dataset and typically has a clearly established goal. In contrast, unsupervised learning is characterised by the absence of predefined objectives; AI learns without specific direction. These two approaches are essential for understanding how AIs can generate other AIs and whether they can do so autonomously.
In supervised learning, we have observed that AIs can indeed learn to create other AIs. This process involves AI being trained by humans with a specific goal in mind. When AI achieves this goal, it is considered successful in its supervised learning. In this sense, we have succeeded in having AI generate new AIs through human direction and training. Following this process raises the question of whether these generated AIs can, in turn, create other AIs or if their capacity is limited to specific tasks designed by humans.
While we have seen examples of AIs creating other AIs, it is essential to clarify that these generated AIs are, essentially, highly specialised algorithms. Unlike what is often depicted in science fiction, where AI is portrayed as an entity with consciousness, morality, and astonishing abilities, the reality of AI in the present world is more pragmatic. Modern AIs are advanced and highly efficient algorithms specialised in specific tasks. Any AI they generate will also be an algorithm designed for a particular function. An example of this is Google’s “AutoML.”
Google’s AutoML – AI Offspring
Google, one of the leading companies in AI development, has made significant advancements in this field with its “AutoML” technology. The creation of AutoML stemmed from a challenge that Google faced: the demand for time and human resources in building machine learning algorithms. From the outset, Google had a clear objective: to create an AI capable of assisting in constructing other AIs, specifically machine learning models. AutoML was designed with this purpose in mind, and through training, it learned to develop machine learning algorithms that are equally effective as those created by humans.
What makes AutoML even more impressive is that the “AI offspring” it generates are often more accurate than those developed by humans. Additionally, as part of the work is performed by the primary AI rather than human programmers, the process becomes less labour-intensive.
AutoML in Action, NASNet
A notable example of AutoML in action is the development of NASNet, an AI offspring explicitly created for object recognition. NASNet proved 1.2% more effective in its task than any other existing system. While these achievements are impressive and demonstrate the potential of AI to generate highly specialised AI offspring, we are still far from achieving the goal of matching or surpassing human intelligence.
Despite their utility in specific tasks, AI offspring like NASNet are far from emulating human intelligence in its entirety. Google’s approach through AutoML is promising and is available in the current market for training customised machine learning models.
Can AI Create Another AI without Human Direction?
Unsupervised learning in AI has witnessed remarkable advancements in recent years, suggesting the possibility of AI learning to generate other AIs autonomously. One of the most intriguing examples of this advancement is the “Paired Open-Ended Trailblazer” (POET) algorithm. POET is a system developed by Uber’s AI division and is characterised by its open-ended approach. Instead of having a specific goal, POET continuously generates new environments and challenges for AI agents to overcome.
In this approach, there is no predefined objective, and AI agents learn to solve problems posed by ongoing obstacles. Once a problem is solved, a new one is created. This system could function indefinitely, constantly generating new problems and, thus, new solutions. This opens the possibility that, in the future, one of these solutions may be the autonomous creation of a new AI.
The Future of AI: Quantum Computing and Autonomous AI Creation
Quantum computing has the potential to address computational and energy challenges in AI. Therefore, quantum computing emerges as a critical element in the evolution of artificial intelligence. While it is still developing, the relationship between quantum computing and AI holds promise. This connection could answer the question: When will AI be able to create other AIs independently?
Until now, the creation of AI by AI has been a challenge, primarily requiring human oversight. However, advancements in unsupervised learning and quantum computing could radically change this scenario.
The autonomy of AI to create other AIs is continually evolving. Although we still need to guide this process, the potential for a radical transformation in technology and artificial intelligence is undeniable. The synergy between quantum computing and AI could pave the way for a new chapter in the history of technology.
Ripple has issued a warning regarding the rise of scam accounts impersonating the company on Telegram. The company clarified that it does not have an official Telegram channel, urging users to stay vigilant against potential fraud.
Summary
- Ripple confirms it has no official Telegram channel for support.
- Scammers use Ripple branding and CEO photos to deceive users.
- XRP Ledger grows, with over 7.7 million holders amid rising scams.
RippleX, a division of Ripple, recently warned the public about an increase in impersonation accounts on social media platforms like Telegram. Fraudsters have been creating accounts pretending to be Ripple recruiters, customer support representatives, or other employees. These scammers often use the company’s branding and images, including pictures of Ripple CEO Brad Garlinghouse, to deceive potential victims.
Ripple emphasized that it does not conduct business through unofficial channels such as Telegram. The company assured its community that it will never contact users directly to offer support, request personal information, or ask for payments.
“Any account claiming to be an official Ripple Telegram is not legitimate,” Ripple stated in a tweet.
How Scammers Operate
Fraudsters have been using various methods to gain trust and trick individuals into sending money. Scammers frequently post fake cryptocurrency giveaways that appear to be associated with Ripple. These fraudulent offers may use genuine videos from Ripple’s media interviews or public events, only to link victims to fake websites or crypto wallet addresses.
Ripple advised the XRP community to remain cautious when approached through unofficial communication channels. The company recommended that users verify any offers or communications by checking through official Ripple platforms.
Despite the rise in scams, Ripple’s XRP Ledger continues to grow, with an increasing number of wallets holding XRP. According to recent reports, the XRP Ledger now has over 7.7 million holders, with a significant rise in wallet addresses. As adoption of XRP grows, the company’s efforts to combat fraud are becoming more important to ensure the safety of its users.
The online gambling industry has settled into a pattern over the past few years. A handful of large operators control most of the market, players pick from similar-looking products, and the biggest innovations tend to be minor updates to existing features. But every now and then a new platform arrives that forces a different kind of conversation. ZunaBet, which launched in 2026, is doing exactly that. Comparing it to a giant like FanDuel reveals just how much distance has opened up between what traditional operators offer and what a new generation of crypto-focused platforms are putting together.
FanDuel: The Household Name
FanDuel needs little introduction. It began as a daily fantasy sports company in 2009 and became one of the dominant forces in US sports betting after federal law changed in 2018. Today it operates under Flutter Entertainment, one of the world’s largest gambling groups, and holds licenses across numerous US states.
Sports betting is the engine of FanDuel’s business. The platform covers every major American sport along with international leagues, offering competitive lines and a polished mobile app that consistently ranks among the best available. FanDuel also runs an online casino product in jurisdictions where it is permitted, providing slots, table games, and some live dealer content. The casino side is functional but clearly plays a supporting role to the sportsbook.
Payments run entirely through traditional channels. Bank transfers, debit cards, PayPal, Venmo, and similar options make up the deposit and withdrawal methods. Processing times vary — deposits are usually quick, but withdrawals can take anywhere from same-day to several business days depending on the method chosen.
FanDuel’s promotional strategy leans heavily on sportsbook offers. New users typically receive some form of bonus bet or protected first wager. Casino promotions exist but tend to be smaller in scope. The loyalty program ties into Flutter’s wider rewards ecosystem, converting wagering activity into redeemable points. It functions as expected without doing anything to differentiate itself from the rewards programs at DraftKings, BetMGM, or Caesars.
FanDuel is a polished, heavily regulated product that delivers a reliable sports betting experience for the US market. What it is not is a platform that feels like it is pushing the industry forward.
ZunaBet: Purpose-Built for What Comes Next
ZunaBet entered the market in 2026 under the ownership of Strathvale Group Ltd. It is licensed in Anjouan and was created by a team carrying more than 20 years of gambling industry experience. The platform was not adapted from an existing product or pivoted from another business model. It was designed from a blank page as a crypto-native casino and sportsbook.
The casino library is staggering in its scope. ZunaBet carries more than 11,000 games sourced from 63 providers. Pragmatic Play, Evolution, Hacksaw Gaming, Yggdrasil, and BGaming headline the list, with dozens of additional studios filling out a catalog that covers everything from slots to RNG table games to live dealer rooms. To put the scale in perspective, FanDuel’s casino product — in the states where it operates one — offers a fraction of that number. ZunaBet’s game selection puts it in the top tier of crypto casinos globally.
The sportsbook stands on its own as a complete betting product. Coverage includes football, basketball, tennis, NHL, combat sports, and virtual sports. The esports section is particularly robust, with active markets for CS2, Dota 2, League of Legends, and Valorant. FanDuel has deeper integration with US sports leagues and markets, but ZunaBet counters with broader global coverage and a level of esports depth that most traditional operators have not yet matched.
On the payment side, the two platforms could not be further apart. FanDuel is locked into fiat currency and traditional banking rails. ZunaBet supports more than 20 cryptocurrencies — BTC, ETH, USDT on multiple chains, SOL, DOGE, ADA, XRP, and others. No processing fees. Fast withdrawals. No need for players to involve a bank or payment processor at any stage. For anyone who holds crypto and has ever waited days for a fiat withdrawal to clear, the difference in experience is immediately obvious.
New players at ZunaBet get a welcome package totaling up to $5,000 in matched deposits plus 75 free spins. That breaks down to a 100% match up to $2,000 with 25 spins on the first deposit, 50% up to $1,500 with 25 spins on the second, and 100% up to $1,500 with 25 spins on the third. Measured against FanDuel’s typical introductory offers, particularly on the casino side, it is a considerably more generous starting point.
The platform itself is built on HTML5 with a clean dark-themed interface that loads quickly and scales across devices. Dedicated apps cover iOS, Android, Windows, and MacOS. Support is available via live chat at any hour.
Two Completely Different Loyalty Philosophies
FanDuel rewards players through its integrated points program. Wagering earns points, points can be exchanged for bonus bets or site credit, and tiered status provides modest upgrades to the overall package. It is a system designed primarily for sportsbook users and mirrors what every other major US operator does. There is nothing wrong with it, but there is also nothing about it that makes a player feel particularly valued or engaged beyond the transactional basics.
ZunaBet approached loyalty as an opportunity to do something players would actually care about. The program revolves around dragon evolution, featuring a mascot named Zuno and six progression tiers — Squire, Warden, Champion, Divine, Knight, and Ultimate. Rakeback begins at 1% for new players and increases all the way to 20% at the highest level. Additional unlocks include free spins scaling up to 1,000, VIP club membership, and double wheel spins.
The structure pulls from game design rather than traditional casino reward logic. Progression is visible, milestones are clearly defined, and each tier feels like a genuine achievement rather than just an arbitrary label attached to a spending threshold. For players who have spent time in gaming environments where leveling up and unlocking rewards is part of the core experience, this kind of system feels natural and motivating. It makes the loyalty program part of the entertainment rather than a background process most players forget about.
The Fiat vs Crypto Divide
This comparison highlights something bigger than just two platforms. It exposes the growing divide between how traditional gambling operators handle money and what crypto-native players actually want.
FanDuel, along with DraftKings, Caesars, BetMGM, and other mainstream brands, was built on top of legacy payment systems. Credit card processors, bank transfers, e-wallets — all of these involve intermediaries that add time, cost, and complexity to every transaction. A player who wins on a Sunday night might not see those funds in their bank account until Wednesday. That has been the accepted reality for years, but it is not a reality that crypto users are willing to accept when alternatives exist.
ZunaBet eliminates that entire layer. Crypto deposits confirm in minutes. Withdrawals process without sitting in a queue. There are no conversion fees, no processing charges from the platform, and no third-party payment company sitting between the player and their money. The system works the way crypto is supposed to work — fast, direct, and without unnecessary intermediaries.
This is not just a convenience difference. It reflects a fundamentally different philosophy about how a gambling platform should relate to its players’ money. FanDuel operates within a system where delays and fees are built into the infrastructure. ZunaBet operates in a system where they have been engineered out of it entirely.
What This Comparison Actually Tells Us
FanDuel is a dominant force in legal US sports betting and that is unlikely to change in the near term. It has the brand, the licenses, the partnerships, and the user base to maintain that position. For players who want a regulated US sportsbook with a familiar interface and mainstream payment methods, FanDuel delivers exactly that.
But the market does not stand still. The number of players holding and using cryptocurrency continues to grow. Expectations around transaction speed and cost are shifting. A new generation of gamblers is arriving with preferences shaped more by gaming culture than by traditional casino culture. These players want bigger game libraries, faster payments, more engaging reward systems, and platforms that feel built for how they live now rather than how the industry operated five years ago.
ZunaBet was clearly designed with these players in mind. Over 11,000 games, 20+ cryptocurrencies, no processing fees, a $5,000 welcome package, a full sportsbook with serious esports coverage, apps on every major platform, and a loyalty program that borrows from gaming rather than copying from other casinos. It is a product that reads like a direct response to every limitation that traditional operators have been slow to address.
ZunaBet is still in its early days. FanDuel has years of operational proof behind it. But if the question is whether the market is shifting, the answer is visible in what ZunaBet has built. The future of online gambling is not going to look like a slightly updated version of the past. It is going to look a lot more like what ZunaBet is already offering.
Strategy (MSTR), already the world’s biggest corporate holder of bitcoin , is on track to record its second-largest quarterly accumulation, continuing its aggressive treasury expansion even as the cryptocurrency’s price sank 20%.
Since January, the company has bought 89,618 BTC, bringing its total holdings to 761,068 BTC. With two Mondays still left for potential purchase announcements this quarter, that number could grow even further.
The only time Strategy has bought more bitcoin was fourth-quarter 2024, when it added 194,180 BTC. That November alone accounted for three of the company’s five largest purchases, with Strategy buying 27,200 BTC, 51,780 BTC, and 55,500 BTC in quick succession as the price surged to $100,000 from $70,000 following President Donald Trump’s second election victory.
In contrast, the past three months have seen bitcoin’s price slump to a level that is now more than 40% below October’s record high $126,000. Strategy’s common stock has dropped 15%.
Recent purchases have been partly funded by sales of the company’s perpetual preferred offering, Stretch (STRC), which accounted for up to 15,000 BTC over the past two weeks. However, as the STRC price failed to reach its $100 par value this week, the company has been unable to utilise the program for now.
Strategy’s accumulation is not just price-dependent. It is driven by capital availability.
A 66-year-old Hong Kong retiree lost 6.6 million Hong Kong dollars (roughly $840,000) in a string of three related crypto investment scams after repeatedly trusting self-proclaimed “virtual currency experts” who reached out via WhatsApp, according to Hong Kong police’s CyberDefender unit.
In a March 20 Facebook post, police said the victim was first approached in September 2025 by a scammer who cold messaged, claiming to be a “virtual currency investment expert” and promising steady gains if the victim followed his advice. The retiree then transferred $180,000 and deposited crypto into a wallet the scammer controlled, only to watch him disappear, prompting the filing of a police report.
The case shows how fraudsters can recycle the same victim through successive schemes that start with “guaranteed profit” pitches and escalate into offers to recover funds that have already been stolen.
“Life has no take two; but scams can have take three,” the CyberDefender team wrote, warning that genuine professionals do not rely on random outreach and that phrases such as “guaranteed returns” and “inside information” are classic red flags.
Related: How US investigators traced $61M in crypto tied to romance scams across wallets
Triple “crypto expert” scam drains retiree’s savings
The retiree then transferred $180,000 and deposited crypto into a wallet the scammer controlled, only to watch him disappear, prompting the victim to file a police report.
Unwilling to accept the loss, the victim later searched online for another “crypto expert” who claimed he could help recover the missing funds, but then demanded $75,000 as a security deposit. After the victim paid, that expert also vanished.
In January, a third supposed specialist messaged the retiree on WhatsApp offering to reclaim both prior losses if the victim bought $585,000 in crypto and sent it to a specified address. Once the victim complied, that scammer disappeared as well, bringing the total losses over roughly six months to approximately $840,000.
Incident falls amid rising Web3 fraud
The case lands against a broader backdrop of mounting crypto-related crime. Web3 platforms saw about $3.95 billion in losses in 2025, with state-linked hackers and weak key security driving much of the damage, according to security firm Hacken.
Authorities worldwide have also flagged new waves of phishing and investment fraud, from the FBI’s recent warning over fake FBI tokens on Tron to India’s GainBitcoin probe and US efforts to forfeit $3.4 million in Tether tied to a multi-state investment scam.
Magazine: Influencers shilling memecoin scams face severe legal consequences
Bitcoin’s mining difficulty shifted lower once more, declining by about 7.7% in the latest retarget to 133.79 trillion at block 941,472, according to CoinWarz data. The move follows a mid-March dip that pulled the metric from roughly 148 trillion to the current level, marking the sharpest drop since February. A lower difficulty means less computational work is required to mine a given block, effectively boosting revenue per unit of hash power for operators that keep running.
The adjustment came on the heels of slower-than-target block production over the previous 2,016 blocks. CloverPool’s explorer data show average block times near 12 minutes 36 seconds—well above Bitcoin’s 10-minute target—prompting the protocol to recalibrate downward to maintain steady issuance.
February’s landscape also featured a notable disruption: weather-related outages in the United States temporarily knocked several large mining facilities offline, triggering a sharp drop in difficulty. As power conditions normalized and hashrate returned, the metric rebounded by roughly 15% in subsequent weeks, underscoring the sensitivity of the network to regional outages and the geographic concentration of mining capacity.
Bitcoin’s difficulty metric measures how hard it is to find a valid hash for the next block. It auto-adjusts to keep block production close to one every 10 minutes; rising hashpower pushes difficulty higher to prevent blocks from being mined too quickly, while a retreat in hashrate lowers the target to preserve issuance cadence.
Related: Cango reports $285M Q4 loss as Bitcoin mining costs surge in 2025
The market consensus around the near-term difficulty path remains conditional on how quickly the next 10-minute cadence can resume as hashrate shifts with weather, power prices and utilization of mining hardware across regions. The next difficulty adjustment is currently projected for April 3, subject to block-by-block changes.
Key takeaways
- March 20 adjustment: Bitcoin mining difficulty fell about 7.7% to 133.79 trillion at block 941,472, marking the steepest drop since February and reflecting a softer recent hash rate.
- Block-time pressure: Average block times around 12 minutes 36 seconds, well above the 10-minute target, catalyzed the downward recalibration to keep issuance stable.
- Weather-driven volatility: February’s drop followed US weather disruptions that temporarily sidelined major facilities, with a roughly 15% rebound as power conditions normalized.
- Strategic shifts among miners: In response to tighter margins and power costs, several operators are moving toward AI and high-performance computing workloads to diversify revenue streams beyond pure BTC mining.
Miner strategy shifts in a power-cost environment
The latest difficulty reset arrives at a moment when a subset of publicly listed miners is broadening its focus beyond traditional Bitcoin mining. Industry observers note that AI workloads and HPC infrastructure offer a potential counterbalance to volatile crypto earnings, leveraging existing data-center footprints and power networks to monetize idle capacity without relying exclusively on block rewards.
Among the players cited in market discourse, Core Scientific, Marathon Digital Holdings (MARA), Hut 8, and Cipher Mining have steered capacity toward AI-oriented deployments or high-performance computing. The trend aligns with a broader re-evaluation of capital expenditure and capacity utilization as power prices squeeze margins and competition for electricity intensifies between compute-intensive sectors.
Additionally, Bitdeer has moved to shrink its treasury exposure. The company disclosed it liquidated 943 BTC from reserves in February and, in its latest weekly update on March 21, confirmed that its BTC holdings remained at zero. Such treasury management moves highlight a broader investor question: how miners balance balance sheets against cyclical earnings and shifting demand for computing power.
Proponents of the AI pivot argue that the overlap between data-center capacity and AI workloads offers a path to steadier returns in environments where BTC mining margins can swing with electricity costs and network difficulty. Critics contend that AI demand may also be volatile and energy-intensive, potentially creating its own cycle of capacity constraints and price pressures.
Industry commentary has also touched on resilience questions for Bitcoin itself. Some observers have framed AI as the newest competing demand for electricity, even as proponents stress the enduring value of Bitcoin’s decentralized security model. The debate underscores a broader strategic tension facing miners: diversify beyond a single revenue line or double down on core hash-power economics during periods of elevated energy costs.
Looking ahead, investors and operators will watch how the next rounds of capacity expansion, power pricing, and regulatory developments influence both the profitability of existing mines and the viability of AI-centric data-center deployments. The ongoing swing in hashrate and difficulty will continue to interact with these strategic choices, shaping the industry’s trajectory through the rest of the year.
As the network navigates these crosscurrents, the immediate question for market participants is what the April 3 adjustment will reveal about the balance of supply and demand in the global mining ecosystem. For readers tracking risk and opportunity, the evolving demand backdrop for AI workloads, the pace of capacity reallocation, and potential regulatory developments in key mining hubs remain critical to watch in the near term.
Readers should stay tuned for the forthcoming data on next-block production and power-market dynamics, which will cast further light on whether miners can sustain growth amid rising energy costs and a shifting compute landscape.
The crypto market saw several important developments today, including a warning from Hong Kong authorities about cryptocurrency scams, a new filing from Grayscale for a crypto-based ETF, and progress on the CLARITY Act in the U.S. Here’s a quick overview of the major events.
Summary
- Hong Kong senior lost HK$6.6M in three crypto scams involving fake experts.
- Grayscale files for HYPE ETF, offering exposure to Hyperliquid’s token.
- US lawmakers near agreement to regulate stablecoin yield to protect banks.
Hong Kong police warn after senior man falls victim to scams
Hong Kong’s Police Cyber Crime Bureau issued a warning today after a 66-year-old retired man lost HK$6.6 million to three separate cryptocurrency scams. According to reports, the elderly victim was first contacted in September 2025 by a fraudster claiming to be a cryptocurrency expert. The scammer convinced the victim to invest, promising guaranteed profits. The man transferred HK$1.4 million to the fraudster, only to realize later that he had been tricked.
Undeterred, the victim sought help from another fraudster posing as an expert to recover his losses. However, after paying a deposit of 600,000 yuan, the second fraudster also disappeared. In January of this year, the victim was once again approached by a scammer claiming to recover the previous losses. This time, the fraudster instructed the victim to purchase cryptocurrency worth 4.6 million yuan, which the victim did. Once again, the scammer vanished, leaving the man without his entire life savings.
Grayscale files for HYPE ETF linked to Hyperliquid token
In other news, Grayscale filed with the U.S. Securities and Exchange Commission to launch an exchange-traded fund (ETF) tied to Hyperliquid’s native token, HYPE. The proposed Grayscale HYPE ETF, if approved, would allow investors to gain exposure to the token’s price movement without holding the token directly.
Hyperliquid is a blockchain platform focused on decentralized perpetual futures trading. The proposed ETF would initially track the price of HYPE, with the potential for staking to be added later. Grayscale’s move adds to a growing list of firms exploring investment products tied to newer digital assets like HYPE, as interest in crypto ETFs continues to expand beyond Bitcoin and Ethereum.
U.S. lawmakers work on stablecoin yield agreement
Meanwhile, in the United States, progress on the CLARITY Act is moving forward. Reports suggest that lawmakers are close to a tentative agreement on stablecoin yield, a key issue that has slowed the progress of the cryptocurrency market structure bill earlier this year.
The proposed agreement would address concerns over stablecoin yield and its potential impact on bank deposits. If passed, the legislation could regulate how stablecoin issuers offer yield to their holders. The deal aims to protect innovation while limiting the risk of deposit flight from the banking system. It could be a significant step forward in regulating digital assets and stabilizing the U.S. crypto market.
Bitcoin’s mining difficulty fell by around 7.7% at the latest adjustment on March 20 to 133.79 trillion at block 941,472, the sharpest drop since February, according to CoinWarz data.
The latest move takes difficulty down from around 145 trillion in mid-March and roughly 148 trillion at the start of the year. A lower difficulty means it takes less computational work to earn the same block reward, slightly improving revenue per unit of hashrate for firms that stay online.
The adjustment followed slower-than-target block production over the prior 2,016 blocks. CloverPool data showed average block times at about 12 minutes 36 seconds, well above Bitcoin’s 10-minute target, forcing the network to recalibrate lower.
In February, difficulty dropped sharply after weather-related disruptions in the United States temporarily knocked large American mining facilities offline, and it later rebounded by about 15% as hashrate returned to the network once power conditions normalized.
Bitcoin (BTC) difficulty measures how hard it is for miners to find a valid hash for the next block and is automatically adjusted to keep issuance steady at one block every 10 minutes.
When more computing power, or hashrate, joins the network, difficulty rises to prevent blocks from being mined too quickly, while a decline in hashrate triggers a lower difficulty, making it easier for remaining miners to earn rewards.
Related: Cango reports $285M Q4 loss as Bitcoin mining costs surge in 2025
The next difficulty adjustment is currently estimated for April 3, though that projection changes with each new block.
Miners pivot to AI as power costs bite
The difficulty reset also comes as several listed miners push further into AI and high-performance computing infrastructure in search of steadier returns on power and data-center capacity.
Last week, crypto trader Ran Neuner argued AI had become Bitcoin mining’s biggest competitor as both industries compete for electricity, even going as far as to say that “AI has killed Bitcoin forever.”
Bitcoin miners such as Core Scientific, MARA Holdings, Hut 8 and Cipher Mining have begun reallocating capacity or pivoting toward AI workloads, while some operators have reduced hashrate or shut down less efficient rigs as profitability tightens.
On Feb 21, Bitdeer liquidated 943 BTC from reserves and sold newly mined coins, cutting corporate holdings to zero. In its latest weekly update on March 21, it confirmed that its BTC holdings remained at zero.
Big questions: Would Bitcoin survive a 10-year power outage?
Former FTX CEO Sam Bankman-Fried, who is currently serving a 25-year sentence for fraud, has renewed public praise of U.S. President Donald Trump, adding to speculation that he hopes to secure a pardon.
In a recent post on X, written through a proxy using prison-approved communications, Bankman-Fried backed Trump’s decision to launch strikes against Iran. He framed the move as necessary to counter nuclear risk and claimed the operation had sharply reduced Iran’s military capacity.
The comments mark his latest in a string of statements supportive of the U.S. president. In earlier posts, he pointed to lower gas prices under Trump than in the Biden era and in other countries. He also credited Trump with “saving” the SEC by replacing former chair Gary Gensler with Paul Atkins, arguing the shift eased pressure on crypto firms and reduced inter-agency conflict.
The tone has drawn attention, given Bankman-Fried’s legal position. Presidential pardons have historically extended to financial crimes, and Trump has shown a willingness to grant clemency in high-profile cases. Ross Ulbricht, who operated a digital black market platform called Silk Road, was sentenced to life in prison without the possibility of parole in 2015 before Trump freed him shortly after being sworn in in 2025. For Bankman-Fried, whose conviction stemmed from one of the largest financial collapses in crypto history, public alignment with the president may serve a clear purpose.
His outreach comes as the remnants of his former empire continue to unwind. Earlier this week, the FTX Recovery Trust said it will distribute about $2.2 billion to creditors as part of an ongoing Chapter 11 process, pushing recovery rates close to full repayment for many claim classes.
Still, the damage from FTX’s collapse runs deep. Millions of customers lost access to funds in 2022, and the event shook trust in crypto markets. Prices fell, firms failed, and regulators stepped in with tighter scrutiny. The case remains a reference point for risk in the industry.
Bankman-Fried’s praise of Trump’s Iran policy lands as that decision faces growing criticism, with some warning the conflict could strain public finances and disrupt global oil supply, as well as concerns about inflation and higher costs for households and businesses.
For now, Bankman-Fried remains behind bars, communicating through intermediaries while his former company repays creditors. His lawyers filed a motion for a new trial in February, which the government opposed. His public messaging, however, suggests he is trying to shape an outcome beyond the courtroom.
The Hong Kong Police Cyber Crime Bureau has issued a warning after a 66-year-old retired man fell victim to three separate cryptocurrency scams.
Summary
- A Hong Kong senior loses HK$6.6M in three separate cryptocurrency scams.
- Police warn against transferring money or crypto to strangers to avoid fraud.
- Fraudsters impersonating experts tricked the victim into losing all his savings.
In total, the elderly victim lost HK$6.6 million after being misled by fraudsters posing as cryptocurrency experts, according to local reports.
In September 2025, the victim received a WhatsApp message from a fraudster claiming to be a cryptocurrency investment expert. The scammer promised guaranteed profits and offered to teach the victim how to invest in cryptocurrencies.
Trusting the individual, the elderly man handed over HK$1.4 million. After the cryptocurrency was deposited into the fraudster’s account, the scammer vanished. Realizing he had been deceived, the victim reported the incident to the police.
Not willing to give up, the victim went online to seek help from another “cryptocurrency expert” to recover his losses. The new fraudster reassured the victim that recovery was possible, but demanded 600,000 yuan as a deposit. Once the victim transferred the money, the so-called expert disappeared as well, and the victim realized he had been scammed a second time.
In January of this year, the victim was contacted by another fraudster claiming to be able to recover his previous losses. This time, the fraudster insisted that the victim purchase 4.6 million yuan worth of cryptocurrency and deposit it into a designated account. As with the previous scams, the fraudster disappeared after the victim complied. At this point, the elderly man had lost his entire life savings to scammers.
Police Warning and Advice
The police are urging citizens to be cautious and avoid transferring money or cryptocurrency to strangers. They emphasize that no one can guarantee to recover losses, and any offer that promises guaranteed returns or insider information is likely a scam. If someone offers to help recover funds after a previous scam, it is likely part of a serial fraud scheme.
Additionally, the FBI has issued a warning about fake tokens on the Tron blockchain impersonating the agency. Scammers are using these fake tokens to trick users into providing personal information under the false pretense that their wallets are under investigation.
Key takeaways
On-chain activity and the ETH exposure narrative
ETF flows, price backdrop, and what they signal
Bitmine’s conviction, DeMark signals, and the long arc of ETH
Sarah Michelle Gellar breaks silence on Nicholas Brendon's death with heart-wrenching “Buffy ”quote
Newscast – The Week: The War In Iran Escalates
Strive Asset Management: Sustained Dip Reflects Flawed Bitcoin As A Treasury Push
-
Tech6 days agoYour Legally Registered ‘Motorcycle’ Might Not Count Under Proposed US Law
-
Politics20 hours agoJenni Murray, Long-Serving Woman’s Hour Presenter, Dies Aged 75
-
Tech4 days agoAre Split Spacebars the Next Big Gaming Keyboard Trend?
-
Fashion21 hours agoWeekend Open Thread: Adidas – Corporette.com
-
Sports7 days ago
Why Duke and Michigan Are Dead Even Entering Selection Sunday
-
Business6 days agoSearch for Savannah Guthrie’s Mother Enters Seventh Week with No Arrests
-
Crypto World7 days agoCoinbase and Bybit in Investment Talks: Could Bybit Finally Enter the US Crypto Market?
-
News Videos3 days agoRBA board divided on rate cut, unusually buoyant share market | Finance Report | ABC NEWS
-
Crypto World15 hours agoNIO (NIO) Stock Plunges 6.5% as Shelf Registration Sparks Dilution Worries
-
Business5 days agoAustralian shares drop as Iran war enters third week
-
Crypto World5 days agoCrypto Lender BlockFills Enters Chapter 11 with Up to $500M in Liabilities
-
Politics3 days agoThe House | The new register to protect children from their abusers shows Parliament at its best
-
Fashion5 days ago25 Celebrities with Curly Hair That Are Naturally Beautiful
-
Tech1 day agoinKONBINI Lets You Spend Summer Days Behind the Register
-
Crypto World3 days agoCanada’s FINTRAC revokes registrations of 23 crypto MSBs in AML crackdown
-
Politics4 days agoReal-time pollution monitoring calls after boy nearly dies
-
Crypto World7 days agoCrypto Losses Drop 87% in February, But Hackers Are Now Targeting People, Not Code
-
NewsBeat3 days agoResidents in North Lanarkshire reminded to register to vote in Scottish Parliament Election
-
Business5 days agoMeta planning major layoffs as AI spending and automation reshape workforce
-
News Videos3 days agoPARLIAMENT OF MALAWI – PAC MEETING WITH REGISTRAR OF FINANCIAL ON AMARYLLIS HOTEL – INQUIRY LIVE
You must be logged in to post a comment Login