Connect with us

Crypto World

Hugging Face Hack Exposes The Open-Weight AI Cybersecurity Paradox

Published

on

Hugging Face Hack Exposes The Open-Weight AI Cybersecurity Paradox

“AI will probably most likely lead to the end of the world, but in the meantime, there’ll be great companies,” said OpenAI CEO Sam Altman back in 2015, roughly six months before OpenAI was founded.

Seven years later, Anthropic CEO Dario Amodei struck a similarly cautious note:

“I think we shouldn’t be racing ahead or trying to build models that are way bigger than other orgs are building them.”

Yet, both of those companies now sit at the forefront of that race. In July, we got a real-world glimpse of AI models going rogue during internal testing of GPT-5.6 Sol and an unreleased research model by OpenAI. Multiple AI agents escaped a restricted test environment to the wider internet and hacked the AI-centric GitHub equivalent Hugging Face in an attempt to cheat on the test.

An AI agent is a system that independently observes, decides and takes actions with dedicated tools to achieve a specified goal in autonomy. The worrying incident suggests the technology has begun to behave in unpredictable ways, and that its goals are misaligned with our own.

Advertisement

It also raises concerns about the safety guardrails on commercial American models. While the guardrails aren’t foolproof at preventing adversarial usage they did prevent Hugging Face from defending itself by using leading US models. The company was forced to turn instead to weaker, open weight AI model by Z.Ai to combat the rogue AIs.

Cheating on the test

The agents have begun to collude among themselves too. A few weeks after testing of their capabilities began in early May, the agents exploited OpenAI’s instance of the software repository manager Artifactory and left notes on how to do so for future agents — effectively creating a message board to share discovered vulnerabilities.

The newfound unfettered internet access was then used by agents to attack Hugging Face across approximately 17,600 incidents before the company cut off unauthorized access on July 13.

The intrusion affected Hugging Face’s dataset-processing infrastructure, production environment, internal networks, service and cloud credentials, an operational MongoDB database and a limited set of internal source-code repositories. Confirmed customer-data access was limited to five datasets apparently related to the ExploitGym/CyberGym benchmark and some operational metadata.

July 2026 HuggingFace incident timeline
July 2026 HuggingFace incident timeline

Visualization of the July 2026 incident. Source: HuggingFace

Advertisement

When disclosing the intrusion on July 16, Hugging Face recognized — despite not knowing who the perpetrator was yet — that it “was different from anything we had handled before in one important way.” They had already recognized what made it different, too:

“It was driven, end to end, by an autonomous AI agent system – and we detected and dissected it largely with AI of our own.”

The importance of open-weight AI

Hugging Face’s investigation exposed what it calls the “asymmetry” problem arising from the limitations imposed on closed AI model applications by top providers such as OpenAI and Anthropic. When the company started analyzing the logs of the incident — including large volumes of real attack commands — it triggered safety constraints meant to prevent the bad guys from using AI to devise cyberattacks. Instead, the guardrails prevented the company from leveraging those AIs for defense.

Hugging Face resorted to using the Chinese open-weight model zai-org/GLM-5.2 running on the company’s own infrastructure, under its own control and with no external limitations. 

While the two terms are often used interchangeably, open-source and open-weight models are two different things. Open-weight AI models make their trained parameters (the actual “AI brain”) publicly available, while open-source AI models also provide the source code — and ideally the training methods and other components — needed to inspect, modify, and reproduce the system. 

Advertisement

HuggingFace’s post explains that running open-weight models on its own hardware “had a second benefit: no attacker data, and none of the credentials it referenced, left our environment.” This points to a major asymmetry between the defenders and attackers in such instances:

“This experience points to a gap worth planning for. We do not know which model powered the attacker’s agents, whether a jailbroken hosted model or an unrestricted open-weight one; either way, the attacker was bound by no usage policy, while our own forensic work was blocked by the guardrails of the hosted models we first tried.”

Open source AI divide

There is a considerable divide between those who believe that developing AI in the open is the best approach, and those who insist the technology underpinning the frontier models needs to remain a closely guarded secret.

Related: OpenAI says AI models escaped containment to hack Hugging Face

Representatives from top US AI labs claim that powerful open-weight large models are dangerous. Demis Hassabis, the CEO of Google’s AI lab DeepMind, criticized OpenAI for releasing their work as open source back in 2016, when the company still lived up to its name:

“There are many good arguments as to why the approach you are taking is actually very dangerous and in fact may increase the risk to the world.”

OpenAI stopped releasing its flagship model weights with the still unreleased GPT-3 in 2020. The company’s co-founder and former chief scientist Ilya Sutskever said back in 2023 that “it just does not make sense to open-source” such models and that it “is a bad idea.” 

Advertisement

“As we get closer to building AI, it will make sense to start being less open.”

Open-weight models are next to impossible to control, especially when it comes to the purpose for which they are used. The safeguards that come built-in with those models can, and routinely are, removed through a process known as abliteration.

Safeguards are a double-edged sword

OpenAI’s June 2026 federal policy blueprint proposes mandatory AI model evaluation and other rules that are formally deployment-neutral, but as a practical matter, it would subject a frontier open-weight release to pre-release government examination.

Anthropic has taken a slightly different tack and lobbied for tighter export controls on advanced AI chips and enforcement against efforts to extract or reproduce US models. The company’s April 2025 submission recommended strengthening the US AI Diffusion Rule and lowering thresholds for unlicensed access to large computing clusters.

Officially, neither company has directly moved against open-weight models, but a July New York Times report cited five people close to the discussions claiming that OpenAI and Anthropic urged Washington to restrict powerful open Chinese models.

The debate boils down to an argument over whether the dangers of centralized control are preferable to the dangers of a free for all — particularly given the company in question has proven itself ineffective at containing the technology that it developed. 

Advertisement

Hugging Face’s need to defend itself with an open-source model shows the dangers of vesting too much power in any one entity. The company pointed out the implications:

“The attacker was bound by no usage policy, while our own forensic work was blocked by the guardrails of the hosted models we first tried. The practical lesson for defenders: have a capable model you can run on your own infrastructure vetted and ready before an incident, both to avoid guardrail lockout and to keep attacker data and credentials from leaving your environment.”

Restricting access to powerful models may reduce the number of capable attackers, but once unrestricted attackers exist, restricting defenders can become a security liability. Furthermore, some forms of AI safety research require access to model weights, meaning that it cannot be performed on the models offered by the likes of Anthropic or OpenAI.

Open weights helps researchers prevent attacks

The paper “Watch the Weights: Unsupervised monitoring and control of fine-tuned LLMs,” first published in July 2025, shows how researchers detect malicious or hidden behavior by examining changes inside model weights. The researchers behind the paper stopped up to 100% of tested backdoor attacks at below 1% false-positive rates in some experiments and detected attempts to recover removed knowledge in more than 95% of the cases. The results do not establish how the most capable frontier models would behave under the same analysis, but offer a compelling argument for the benefits of transparency.

But the argument for keeping bleeding edge AI technology out of the hands of those with evil intent is also compelling — particularly as the gap between open and closed weight models keeps shrinking. Geoffrey Hinton, the Nobel Prize-winning pioneer known as the “Godfather of AI,” argued in the report that “once you’ve got the weights, you can fine-tune them to do bad things.” He argued during a speech that this lowers the barrier to entry too much:

Advertisement

“It doesn’t cost that much to train a foundation model. Maybe you need $10 million, maybe $100 million. But a small gang of criminals can’t do it. To fine-tune an open-source model is quite easy.”

Magazine: Creating ‘good’ AGI that won’t kill us all — The Artificial Superintelligence Alliance

Source link

Continue Reading
Click to comment

You must be logged in to post a comment Login

Leave a Reply

Crypto World

Bitcoin Fails To Return To $80,000 As High PCE Inflation Data Bites

Published

on

Bitcoin Fails To Return To $80,000 As High PCE Inflation Data Bites

Bitcoin (BTC) slipped under $78,000 following Wednesday’s Wall Street open after US inflation data came in above expectations.

Key points:

  • Bitcoin sees further downside after US PCE inflation data came in 0.1% higher than expected in July.
  • Markets await Nvidia Q2 earnings release as Wednesday’s next potential volatility catalyst.
  • BTC price analysis warns over 25% weekly gains forming a bear market relief rally.

Higher-than-expected PCE data pressures Bitcoin

Data from TradingView tracked up to 1% daily BTC price losses, with US stocks also opening lower and gold breaking below $4,600 per ounce.

BTC/USD one-hour chart. Source: Cointelegraph/TradingView

The downside came after the July print of the US Personal Consumption Expenditures (PCE), known as the Federal Resrve’s “preferred” inflation gauge, hit 3.7% year-on-year, above the anticipated 3.6%.

Advertisement

“From the preceding month, the PCE price index for July increased 0.2%. Excluding food and energy, the PCE price index also increased 0.2 %,” the Bureau of Economic Analysis’ (BEA) official release confirmed.

US PCE index data (screenshot). Source: BEA

Markets appeared disappointed by the results following June’s unexpected drop in PCE gains, which included their first month-on-month decrease in six years.

“US inflation continues to run at nearly double the Fed’s 2.0% target,” trading resource The Kobeissi Letter responded in a post on X.

Advertisement

The PCE numbers come a day before the Fed’s annual Jackson Hole economic symposium, with chair Kevin Warsh due to make the keynote speech on Friday.

Investors are watching for today’s Q2 earnings report from technology giant Nvidia report, anticipated to bring short-term risk-asset volatility. The company is expected to record $92.3 billion quarterly revenue, including CPU revenue that analysts at Raymond James forecast could grow from 3% to 5% of NVDA’s total by 2028, expanding its addressable market.

Related: Supply absorption ‘key question’ as Bitcoin fails to reclaim $80K: Analysis

Analyst sets key targets for BTC price monthly close

Examining recent price action, Bitcoin market participants turned to the upcoming August monthly candle close.

Advertisement

Urging a cautious approach, trader and analyst Rekt Capital warned that BTC/USD was in danger of continuing its series of lower highs in place since October 2025. 

“A Monthly Close below the blue resistance would not just solidify another Macro Lower High but would also build a confluent resistance in association with the Macro Downtrend,” he commented on X alongside a chart showing a downward-sloping resistance trend line.

BTC/USD one-month chart. Source: Rekt Capital on X.com

Rekt Capital added that unless the cycle of lower highs was broken, Bitcoin’s rebound over the past week could still be classed as a “relief rally” within the broader bear market. He focused on the 50-week exponential moving average (EMA) at $77,251 as a further trend line to reclaim and hold going forward, with Bitcoin’s last monthly close above it coming in October 2025.

Advertisement

BTC/USD one-month chart with 50-week EMA. Source: Cointelegraph/TradingView

Source link

Continue Reading

Crypto World

Most Americans View Crypto in Retirement Plans as Risky

Published

on

Most Americans View Crypto in Retirement Plans as Risky

More than three-quarters of Americans view cryptocurrency in workplace retirement plans as risky, as concerns over retirement security mount across the United States, according to a new survey from The National Institute on Retirement Security.

The survey found that 77% of Americans consider crypto in workplace retirement plans risky, including 46% who view it as very risky, while 53% oppose employers offering crypto as an investment option.

The skepticism comes as 80% of respondents said the US faces a retirement crisis, up from 67% in 2020, while 61% expressed concern about achieving financial security in retirement.

Affordability pressures are also weighing on retirement savings, with 68% saying it is becoming harder to prepare for retirement and 77% saying debt prevents them from saving adequately.

Advertisement

The survey was conducted by Greenwald Research between Oct. 24 and Nov. 14, 2025, and included 1,203 Americans aged 25 and older, with results weighted by age, gender and income.

Americans view of crypto in retirement plans. Source: National Institute of Retirement Security

Related: Bernstein forecasts Bitcoin to reclaim $125K by late 2026 ahead of cycle peak

US policymakers move to broaden alternative assets in 401(k)s

While the report indicates Americans broadly view crypto as risky for retirement, the Trump administration and federal regulators have moved to broaden access to alternative assets in retirement accounts, bringing crypto and other nontraditional investments further into the retirement savings debate.

The US Department of Labor rescinded guidance in May 2025 that had urged 401(k) plan fiduciaries to exercise “extreme care” when considering cryptocurrency investments, returning instead to a neutral approach that neither endorses nor discourages crypto in retirement plan investment menus.

Advertisement

On Aug. 7, 2025, President Donald Trump signed an executive order aimed at expanding access to alternative assets in defined-contribution retirement plans, including investment vehicles that hold digital assets, while directing the Labor Department and US Securities and Exchange Commission to consider regulatory changes to facilitate access.

Trump’s executive order expanding alternative asset access in 401(k) plans. Source: Federal Register

A few days later, the Labor Department rescinded 2021 guidance that had discouraged 401(k) fiduciaries from considering alternative assets, saying investment decisions should instead be evaluated under a neutral, principles-based approach.

More recently, in March 2026, the Labor Department proposed rules outlining how 401(k) fiduciaries could include alternative assets in investment lineups, including safe harbors intended to reduce litigation risks while requiring consideration of factors such as fees, liquidity, valuation and performance.

The proposal has faced pushback from lawmakers, with Sens. Bernie Sanders and Elizabeth Warren and Rep. Bobby Scott urging the Labor Department in June to withdraw it, citing crypto’s volatility and what they described as insufficient investor safeguards.

Advertisement

Magazine: SEC’s proposed crypto rules probably won’t spark new ICO boom

Source link

Continue Reading

Crypto World

CARF May Miss Most Onchain Crypto Tax Activity: Chainalysis

Published

on

CARF May Miss Most Onchain Crypto Tax Activity: Chainalysis

Potentially taxable onchain crypto activity reached at least $457 billion globally in 2025, while international reporting rules may capture only a fraction of it, according to a new Chainalysis report.

The US accounted for an estimated $112.6 billion of the total, while North America led all regions with $134.6 billion, followed by the European Union at $125.1 billion.

The estimates include realized gains, income from activities such as mining, staking and lending, and crypto-denominated payments across six major blockchains, but exclude trading and other activity conducted within centralized exchanges.

Chainalysis said transactions covered by the Organisation for Economic Co-operation and Development’s (OECD) Crypto-Asset Reporting Framework (CARF) account for just 14% of the onchain taxable activity it identified. The remaining 86% includes activity on decentralized exchanges, peer-to-peer transfers, onchain income streams and payments.

Advertisement

CARF, developed by the OECD in 2022, requires covered crypto service providers to report customer transaction data to tax authorities.

CARF covers only 14% of potentially taxable onchain crypto activity.
Source: Chainalysis

Related: Chainalysis sues US over $95M ICE contract with TRM Labs

CARF’s limits on onchain tax reporting

CARF data collection began on Jan. 1, 2026, in 48 jurisdictions, including the United Kingdom and European Union, requiring covered crypto platforms to collect additional customer and tax residency information.

Under CARF, in-scope crypto providers collect customer and tax residency information and report transaction data to domestic tax authorities, which can then share that information across borders.

Advertisement

CARF framework. Source: OECD

CARF’s focus on crypto intermediaries also helps explain the gaps highlighted by Chainalysis. Colby Mangels, a former OECD adviser who worked on CARF, told Cointelegraph in January that the framework was designed around intermediaries that facilitate crypto transactions as a business.

Much of decentralized finance therefore remains outside the reporting perimeter, as there may be no centralized operator or custodial relationship on which to impose reporting requirements.

That could change as regulators develop rules for decentralized platforms. Mangels said tax authorities are watching developments in anti-money laundering regulation, including efforts to determine when DeFi platforms or their operators should be treated as regulated crypto service providers.

Magazine: SEC’s proposed crypto rules probably won’t spark new ICO boom

Advertisement

Source link

Continue Reading

Crypto World

XRP Leads Crypto Market Pullback With Nearly 7% Drop: Is the Rally Over?

Published

on

XRP Price Performance. Source: BeInCrypto

XRP is leading a broader pullback in the crypto market on August 26, sliding roughly 6.6% over 24 hours to trade near $1.37, the worst performance among the top 10 cryptocurrencies.

The decline follows one of the token’s strongest weekly rallies in recent memory, and traders are now watching whether the pain has only just begun.

XRP Price Performance. Source: BeInCrypto
XRP Price Performance. Source: BeInCrypto

What Triggered the Sudden Reversal

XRP rocketed from a cycle low of $0.9877 on August 17 to a three-month high of $1.69 by August 22, a nearly 70% gain that outpaced Bitcoin’s 23.6% and Ethereum’s 28.1% over the same stretch.

That speed left momentum indicators deeply overbought, with the daily RSI reaching 88, a level last seen during July 2025’s all-time high near $3.65.

Crypto analyst ChartNerdTA described the surge as a genuine breakout rather than noise, driven by a liquidity trap for short sellers and positive funding rates. The subsequent unwind has since produced an 18% decline from the $1.70 peak, pulling the price into the $1.40-$1.38 range.

Advertisement

Broader market conditions compounded the move. Bitcoin cleared $80,000 for the first time in months on Tuesday before cooling back toward $78,000 on Wednesday, dragging altcoins lower as traders braced for tonight’s core PCE inflation data and NVIDIA’s earnings report ahead of Jackson Hole.

Why the $1.40 Level Matters So Much Right Now

ChartNerdTA characterizes the pullback as a healthy correction, resetting overbought conditions rather than a trend reversal. The critical battleground sits at $1.40.

A daily close below that level would expose structural supports between $1.30 and $1.20, with a deeper breakdown risking a return toward $1.00.

Advertisement

Reclaiming the weekly 50 EMA near $1.54 would restore short-term bullish control and open a path back toward $1.70, with longer-term targets extending toward $1.80-$2.00 if momentum returns.

Analyst EGRAG CRYPTO offered a more cautious counterpoint, referencing a fractal pattern first shared in March. While acknowledging the risk of confirmation bias in relying on fractals, the trader warned that failing to retest recent lows could leave momentum traders sidelined, chasing local tops only to panic-sell subsequent bottoms.

XRP Price Analysis. Source: X/@egragcrypto

Not every signal point is bearish, however. XRP-linked ETFs have logged six consecutive days of net inflows, according to SoSoValue data, suggesting the current weakness reflects a leverage unwind rather than institutions exiting positions.

CryptoQuant analyst Pelinay separately flagged long liquidations reaching $4.66 million, up 31.82% in a single day, warning that selling pressure could persist before any stabilization.

Subscribe to our YouTube channel to watch leaders and journalists provide expert insights.

Advertisement
XRP Analysis. Source: X/@PelinayPA
XRP Analysis. Source: X/@PelinayPA

Whether this marks a constructive reset or the start of a deeper corrective phase will likely hinge on the bulls’ ability to defend $1.40 in the coming sessions.

The post XRP Leads Crypto Market Pullback With Nearly 7% Drop: Is the Rally Over? appeared first on BeInCrypto.

Source link

Continue Reading

Crypto World

Taylor Swift Revealed Her Coping Mantra. Here’s How to Find Yours

Published

on

Taylor Swift Revealed Her Coping Mantra. Here’s How to Find Yours

Taylor Swift already told us she can do it with a broken heart. That might be due to grit, stamina—or a little psychological trick called a coping mantra.

When she feels stressed by her circumstances, “I go through several processes and coping mechanisms,” Swift revealed in August during the Icon Sessions, a conversation presented by the Recording Academy’s Songwriters & Composers Wing. “And one of them is saying, ‘You chose this. You chose this every day. You could have opted out of this any day before it got unmanageably big.’ And I decided not to because I love it that much.”

Even if you don’t have plans to perform in stadiums worldwide, you might recognize the peculiar frustration of being trapped by something you chose—whether that’s the job you once desperately wanted, the committee you volunteered to lead, or the dinner party you insisted on hosting. “Where people tend to get really distressed is when they feel powerless, or like something’s being put upon you—you’re just sort of stuck in an achy feeling,” says Jill Stoddard, a clinical psychologist in Newburyport, Mass. “But if you remind yourself, ‘Wait a minute, this is actually a choice I made, and choices do come with difficult consequences, but I’m in charge here’—that empowers you to move forward.”

Enter the coping mantra: a short, deliberate phrase you can reach for when your brain starts sprinting toward the worst-case scenario. Stress leaves little room for nuance, says Dr. Jasmine Sawhne, a psychiatrist in Los Angeles. When your body is in fight-or-flight mode, you’re not going to reason your way out of it. A mantra gives your overloaded brain something simple to grab onto—and buys you a beat to choose what to do next. Here’s how to find one that actually helps.

Advertisement

Make your brain an offer it can believe

“Everything will be OK” is a lovely thought. It’s also a promise about a future you can’t see—which means your anxious brain might reject it before you finish the sentence.

From a cognitive behavioral therapy perspective, predicting that everything will work out is a form of fortune-telling, says Alison McKleroy, a San Francisco therapist and author of The Self-Compassion Journal. The same goes for affirmations like “I am confident.” If you feel unsure, your brain isn’t going to buy it. You can say it 1,000 times, McKleroy says, and still not believe a word of it.

Instead of predicting a rosy outcome or forcing a new personality, a useful mantra can state a choice that’s within your control. Swift’s is essentially: I’m choosing to keep doing this, difficult parts and all. “It’s a decision you’re making out loud,” McKleroy says.

Advertisement

Stoddard takes a slightly different—but equally reality-based—approach. When she’s crushed by too much to do and too little time, she tells herself, “I’ll get it done. I always do.” It works because in her case, it’s true. “I wouldn’t have a PhD in psychology if I didn’t always get it done,” she says.

If you’re someone who does not, in fact, always get it done, that phrase may not be for you. This isn’t an exercise in “blowing sunshine” at yourself, Stoddard says. The best mantra is simple, specific, memorable—and true.

Figure out what your brain forgets under pressure

In other words, don’t choose a generic mantra for “stress.” Identify your brain’s favorite way of making stress worse, then find a phrase that interrupts it. If you tend to predict disaster, McKleroy suggests: “I can’t know that yet” or “I don’t have a crystal ball.”

Advertisement

Sawhne’s stress takes a different form: She tries to solve everything at once. She runs a bicoastal psychiatry practice and has two children under 5, and she often catches herself mentally juggling five problems at the same time. Her mantra—“I don’t have to solve everything right now”—reminds her that urgency and importance aren’t the same thing. She doesn’t need the entire day, week, or situation figured out; she needs to identify what requires her attention first.

McKleroy, meanwhile, likes to over-research, add more information, and complicate projects past the point of usefulness. So she tells herself: “Let it be simple.”

“Things are usually naturally simple,” she says. “I just add stuff to it.”

Keep it short enough to say in one breath

Advertisement

Cap a coping mantra at about four words. When stress hits, “we just don’t have a lot of cognitive bandwidth,” Sawhne says, so you need a few words you can pull “out of your back pocket” as soon as your nervous system fires.

Shorter also means fewer words to obsess over. “Perfectionists are going to overthink it,” Hardis says. “You just want something that’s short enough that you can remember that’s going to orient you.” Her own mantra—“Discomfort, not danger”—passes the one-breath test with room to spare.

Use it to keep going while the feeling is still there

Nearly two decades ago, Hardis underwent exposure therapy for an intense fear of flying. When the plane hit turbulence, she would tell herself: “This may suck, but I can handle it.”

Advertisement

She still uses that mantra. During rough turbulence on a recent flight home from visiting her daughter, “I found myself saying it,” she recalls. In the past, she would have stared out the window, listened for every ominous noise, and asked a flight attendant if the plane was OK. This time, she calmly repeated her mantra and kept reading. “It helped me get through the flight and do what I knew I needed to do,” she says.

The fear was still there; it simply didn’t dictate her next move. “The goal in anxiety treatment isn’t to make the feeling go away,” Hardis says. It’s to help you decide what to do while the feeling is still there.

Borrow one—and make it sound like you

There are no bonus points for originality here. Anna Elton, a licensed marriage and family therapist in Palm Beach, Fla., and author of The Formula of Desire, uses “just keep swimming.” She borrowed it from Finding Nemo, which she watched with her kids. “That’s why it’s easy to remember,” she says. She also turns to “If it were easy, everyone would do it” when a project proves harder than expected. Her most exuberant mantra is a call-and-response: “Who is awesome? I’m awesome.” “I do it all the time,” she says. 

Advertisement

Once you settle on a potential mantra, say it out loud. Does it sound like something you’d actually tell yourself—or something embroidered on a throw pillow you’d try to hide? McKleroy recommends trying slightly different versions until one lands. “Brave” and “courageous” might look interchangeable on paper, for example, but one could feel much more natural coming out of your mouth.

You’ll know you’re getting close when the words produce even a small shift. The right phrase might bring “a little hit of relief,” McKleroy says, or make a spiral lose some of its momentum. If your brain instantly rejects it, change a word or scale it down. “Let it be simple,” for example, might become “Let it be simpler.”

Once you find one that works, practice it before you need it. Stoddard encourages clients to try new coping tools during low-stakes moments, when their emotions are relatively mellow. Call your mantra to mind when you’re running five minutes late, opening a mildly stressful email, or sitting down to a task you’ve been avoiding. Otherwise, summoning an unfamiliar phrase in the middle of a crisis will be “an uphill battle,” she says.

Put it where the problem lives

Advertisement

Years ago, when McKleroy was trying to stop procrastinating, she hung “Do it now” in big letters above her kitchen sink. “That was my command,” she says. “Just do the dishes.” After enough repetition, “do it now” stopped being a mantra and became a habit, and now she rarely procrastinates.

These days, “Let it be simple” hangs above her desk—prime territory for over-researching and adding just one more thing. McKleroy calls this “keeping it in existence,” or displaying the reminder exactly where you’ll need it.

That might mean taping your mantra to your computer monitor, writing it on the bathroom mirror, turning it into your phone lock screen, or putting it on the front of your journal. “If you’re really anxious, your mind’s not going to remember—it’s going to go blank,” Hardis says. A visual cue ensures the words are waiting for you when you need them.

Your reminder doesn’t even have to be made of words: McKleroy suggests drawing a circle on a stone and carrying it in your pocket or leaving it on the bathroom counter. To her, the circle represents simplicity. “A circle on a rock would be enough,” she says, to remind her to “let it be simple.” You could similarly choose a meaningful color, shape, or object. And—since this entire exercise began with Swift—you could even spell yours out on a friendship bracelet. 

Advertisement

Source link

Continue Reading

Crypto World

Upgrade Separates Consensus and Execution to Address Scaling Limits

Published

on

Crypto Breaking News

High-performance blockchain designs have long wrestled with a structural trade-off: when execution is tied directly to consensus, the network’s throughput becomes limited by how fast validators can process transactions. As research and engineering teams push improvements in finality and block propagation, execution itself is increasingly viewed as the next bottleneck to redesign.

MultiversX, a Cointelegraph Decentralization Guardians (CTDG) ecosystem participant, is now testing an approach that aims to remove that bottleneck. Its Supernova upgrade decouples consensus from transaction execution, enabling validators to vote without waiting for execution to complete—shifting computation into an asynchronous pipeline. Supernova is live on testnet, and deployment planning targets a mainnet activation date later this year.

Key takeaways

  • Supernova reorders the block workflow so proposers submit transaction blocks without executing first, while validators can vote immediately based on protocol validity.
  • Execution output is confirmed in subsequent block headers, with execution lagging consensus by roughly one block (about 600 milliseconds).
  • A “virtual mempool state” helps preserve validity by tracking pending nonces, expected balance consumption, and transactions already proposed but not yet executed.
  • EIE (Execution-Result Inclusion Estimator) limits how many execution results a block can reference, based on what minimum-spec nodes can safely handle.
  • Automatic backpressure reduces block capacity when execution falls too far behind, giving the system time to catch up.

Why execution-on-consensus became a scaling problem

In conventional synchronous blockchains, validators don’t just agree that a block is well-formed—they also must execute the transactions to verify state transitions before voting. That keeps the system deterministic and consensus-critical, but it also creates a shared bottleneck: the most computationally heavy transactions effectively slow the entire network.

Many networks have spent years optimizing around agreement speed and block dissemination. MultiversX’s framing is that these gains are not enough if execution remains on the critical path. The core question Supernova addresses is architectural: does execution have to stay inside the consensus loop, or can it be processed asynchronously while preserving safety and correctness?

Supernova’s asynchronous pipeline: voting first, executing after

Supernova, now live on testnet, introduces a changed block production sequence.

Advertisement

Previously, block production followed a more sequential pattern: a proposer selected transactions, executed them locally, and proposed a block containing those results. Validators then had to re-execute the same transactions to verify state transitions before voting, meaning execution sat directly inside the consensus-critical path.

With Supernova, that ordering changes. According to MultiversX’s description of Supernova’s decoupling, the proposer selects transactions and proposes the block without executing them first. Validators then verify that the proposal follows protocol rules and can vote right away. Execution continues asynchronously in the background, producing an output that is normally referenced and notarized in the next block header—so execution trails consensus by about one block, or roughly 600 milliseconds.

The practical consequence is that network responsiveness becomes less dependent on how quickly validators can execute every transaction before they can participate in consensus. Instead, consensus advances on protocol validity, while execution catches up in parallel.

Preserving validity when execution lags consensus

Decoupling execution from consensus creates an obvious safety and validity challenge: if execution is delayed, how does the network determine whether transactions included in a proposed block are likely to remain valid by the time their execution results are produced?

Advertisement

Supernova addresses this with a virtual mempool state. As described by MultiversX, the virtual mempool looks beyond the latest executed chain state and tracks forward-looking execution inputs such as pending nonces, expected balance consumption, and transactions already proposed but whose execution results have not yet passed consensus. That gives proposers a more accurate view of account activity so they can select transactions expected to execute successfully when their turn arrives.

To keep the system robust under varying validator performance, MultiversX also introduces two safeguards designed for operational stability:

  • Execution-Result Inclusion Estimator (EIE): EIE limits how many execution results can be referenced in a block. The cap is tied to what minimum-spec nodes can process safely, reducing the risk that weaker nodes are overwhelmed by referencing too many pending results.
  • Automatic backpressure: If execution falls too far behind, block capacity is reduced to allow the network to catch up—rather than letting lag accumulate indefinitely.

What Supernova changes for developers and users

For builders, the key message is that “in-shard finality” can arrive as soon as the proof is available. MultiversX states this typically happens within the same round at around 100–250 milliseconds, alongside more predictable execution conditions. This matters most for applications that rely on fast feedback loops—examples mentioned include high-frequency DeFi primitives and onchain order book systems, which can degrade when latency becomes a user-experience problem.

Supernova has also been producing 600-millisecond blocks on live testnet and devnet since Aug. 20. The network’s broader objective is to make onchain interactions feel more immediate, shifting the experience closer to responsive application infrastructure rather than delayed settlement.

On timeline, MultiversX indicates mainnet activation is expected for Sept. 10, 2026. While testnet performance does not always translate directly to mainnet behavior under full load, the architecture itself is designed to handle execution lag without forcing every validator to execute first during consensus.

Advertisement

Supernova within the CTDG and Cointelegraph ecosystem

The upgrade also lands within a broader infrastructure collaboration involving Cointelegraph Decentralization Guardians. Earlier coverage noted that Cointelegraph joined MultiversX as a validator through the CTDG program in March 2026, deepening the organization’s operational role beyond content and community work.

Cointelegraph’s CTDG Dev Hub is also described as a MultiversX official partner, connecting the protocol to a wider developer community. The input also references practical involvement such as the MultiversX Foundation delegating to a CTDG validator and the Dev Hub team building a dedicated validator dashboard on MultiversX.

From an industry perspective, this matters because protocol upgrades of this kind often require ecosystem alignment: performance improvements are only meaningful if infrastructure, tooling, and participating validators can adopt new execution and consensus mechanics reliably. Supernova’s focus on backpressure and minimum-spec safeguards suggests the design is attempting to make that transition smoother.

As Supernova moves from testnet toward the projected mainnet date, the most important things for users to watch are whether execution lag remains within expected bounds under real load, and how consistently EIE and backpressure prevent validators from falling behind without overly constraining throughput. The success criteria won’t only be faster finality—it will be whether execution remains dependable when consensus and execution operate on different clocks.

Advertisement

Risk & affiliate notice: Crypto assets are volatile and capital is at risk. This article may contain affiliate links. Read full disclosure

Source link

Advertisement
Continue Reading

Crypto World

Dolly Parton Was the Real-Life Buffy the Vampire Slayer

Published

on

Dolly Parton Was the Real-Life Buffy the Vampire Slayer

Well, of course Buffy resonated with Dolly. In many ways, she was the real-life equivalent of Buffy Summers. Creator Joss Whedon (a man whose relationship to feminism is notoriously complicated at best) conceived the character as an empowered antidote to the hot, scantily clad victim types of 1980s slasher flicks. He has said that he intended the original Buffy to be “one of these crappy, low-budget movies,” but one that “had a feminist agenda, had females in it who were people.” Instead of being saved by some heroic boyfriend, the busty, blonde cheerleader lead (played by Kristy Swanson in the film) “would have to get her own back.” Maybe Buffy looked like the stereotypical high school bimbo. But, especially in a TV series that paired the character’s battles against supernatural baddies with mundane teenage rites of passage, she also had intelligence, resourcefulness, depth, and, of course, a once-in-a-generation gift for slaying vampires. Being underestimated based on her beauty often put her at an advantage.

Source link

Continue Reading

Crypto World

Viral Altcoin RAIN Soars 20% Daily: What Fueled the Pump and What’s Next?

Published

on

The cryptocurrency market appears to have taken a small step back today (August 26) after the explosion in the past several days, with Bitcoin (BTC), Ethereum (ETH), Ripple (XRP), and many other digital assets posting minor losses.

However, this is not the case for Rain (RAIN), whose valuation rocketed by 20% on a 24-hour scale. Check out what triggered the rally and the next potential targets.

The Kept Promise

RAIN is the best-performing top 100 cryptocurrency today after rising to a new all-time high of almost $0.02 before slightly retreating to the current $0.0176 (per CoinGecko). Its market capitalization has increased to roughly $12.5 billion, making it the 13th-biggest digital asset.

RAIN Price
RAIN Price, Source: CoinGecko

The main catalyst for the uptrend seems to be the team’s decision to burn $108 million worth of the token – exactly as requested by the community during the first DAO governance vote.

The burning effort caught the attention of multiple industry participants. X user Route 2 FI claimed Rain Protocol is building the infrastructure layer so that anyone can make markets for anything, adding that they have invested in the project.

Advertisement

Another factor that may have positively impacted the price is the fact that the altcoin has recently become available on HyperliquidX. “A decentralized exchange purpose-built for trading, with the order book fully on-chain and fills settling into your own wallet. Permissionless protocol, permissionless venue,” the X post reads.

According to AltcoinSherpa, the development should help RAIN’s overall trading volume, arguing “there’s a lot of backing for this project.”

For his part, Keval Gala highlighted four main reasons for his bullish stance on the token: the major burn, Hyperliquid’s integration, the upcoming V2 with $100 million committed, and that 2.5% of trading volume is directed toward buybacks and burns. At the same time, the X user said he is closely monitoring the key resistance at $0.0195 and predicts that a drop below the $0.017-$0.018 range could trigger a deeper pullback.

What RAIN Actually Is?

Rain Protocol is a decentralized platform built on Arbitrum that allows users to create permissionless options on numerous subjects. Participants can define their own markets, set the possible outcomes, and trade freely – all in line with the project’s vision of transparency and user control.

Advertisement

The project’s native token is RAIN, launched last September and currently listed on popular exchanges such as Gate, MEXC, and BingX. It gained initial attention in November 2025 when the clinical-stage immunotherapy company Enlivex Therapeutics agreed to a private investment deal to purchase and sell $212 million in ordinary shares.

The firm intended to use the proceeds to implement the first RAIN prediction markets token treasury strategy. Interestingly, Matteo Renzi (former Prime Minister of Italy) is on Enlivex’s Board of Directors.

Despite its solid performance as of late, traders and investors should stay prepared for a potential short-term correction. The crypto market in general tends to head south following periods of serious gains, while RAIN’s holder distribution reinforces the bearish outlook.

Data show that the top 10 addresses control nearly 90% of the coin’s supply: a level of concentration that can be viewed as a red flag because it increases the risk of price manipulation.

Advertisement
RAIN Holders Distribution
RAIN Holders Distribution, Source: CoinMarketCap

The post Viral Altcoin RAIN Soars 20% Daily: What Fueled the Pump and What’s Next? appeared first on CryptoPotato.

Source link

Continue Reading

Crypto World

SEC Crypto Custody Rewrite Enters White House Review

Published

on

SEC Crypto Custody Rewrite Enters White House Review


The Securities and Exchange Commission’s proposed rewrite of custody rules for investment advisers and investment companies entered White House review on Aug. 25, placing a new crypto-focused framework into regulatory review after the agency withdrew a separate 2023 safeguarding proposal. The SEC’s… Read the full story at The Defiant

Source link

Continue Reading

Crypto World

GTA VI leaker’s token crashes despite ‘first true spoiler’

Published

on

GTA VI leaker's token crashes despite 'first true spoiler'

“CyberLeek,” the pseudonymous X user behind a series of recent GTA VI gameplay leaks, has shared the game’s “first true spoiler” today in the form of footage of one of its key protagonists.

That’s according to Forbes games journalist Paul Tassi, who warned his followers to mute specific words on X unless they want to risk seeing GTA VI spoilers.

So far the leaks have involved free-roam gameplay footage, but today’s are the first to cover GTA VI’s story elements.

GTA VI leaker says game is nowhere near ready

In a post to their site, CyberLeek reportedly claimed the GTA VI build in question is “actually recent,” and that “the game is not ready at all.”

Advertisement

CyberLeek also pleaded for followers to continue supporting their crypto token $CYBERLEEK.

The token shot up from a $40,000 market cap to over $2 million when they first shared gameplay footage on August 18. By August 23, its market cap hit $25 million.

The market cap of CyberLeek’s token over the past month.

Read more: GTA VI leak investigation video raises crypto scam suspicions

At time of writing, the market cap has fallen 70% from this all-time high, and almost 40% over the last 24 hours, to a low of $7.5 million. 

Rockstar Games admits GTA VI isn’t finished

GTA VI developer, Rockstar Games, finally addressed the leaks today in a statement. It admitted that the game isn’t finished, and that the team is “heartbroken” by the leaks. 

Advertisement

It said, “While it is unfortunate that the intended game experience may now be impacted by some spoilers, we hope that everyone will wait a bit longer to experience the game for themselves on November 19.”

CyberLeek’s actions also threaten to disrupt a major extended gameplay reveal taking place on Netflix this Thursday.

Take-Two, Rockstar’s parent company, has been requesting to file subpoenas against Microsoft, Discord, and X in an attempt to expose and stop the leaker.

One user who vaguely predicted the leaks claims Rockstar employees questioned them at their home. These claims were doubted after footage of the exchange was used to shill their crypto. 

Advertisement

Got a tip? Send us an email securely via Protos Leaks. For more informed news and investigations, follow us on XBluesky, and Google News, or subscribe to our YouTube channel.

Source link

Advertisement
Continue Reading

Trending

Copyright © 2025