Welcome to Nofinity

• Manuel Arios, former CTO of OpenZeppelin, declared on May 26, 2026 that he considers all of DeFi unsafe. He advised friends and family to withdraw funds from blue-chip protocols like MakerDAO and Compound.
• OpenZeppelin secures code worth roughly $250 billion in live protocol value and has conducted over 900 audits. This warning comes from one of the industry's most authoritative voices.

• A defender must find and fix every bug; an attacker only needs one. This imbalance is worsened in DeFi by three factors:
  • Immutable code (no patching after deployment)
  • Public code (anyone can study defenses)
  • Transparent money piles (protocols advertise their values openly)
• A 2025 study shows: attackers turn a profit at around $6,000 exploit value, while defenders face $60,000 for the same coverage – a structural 10x cost disadvantage.

• In December 2025, Anthropic's Scone benchmark showed: frontier AI models (Claude, GPT-5) successfully reproduced 51% of all exploits from 405 real contracts (2020–2025), extracting simulated $550 million.
• For novel contracts (deployed after model knowledge cutoff), 19 of 34 were cracked – $4.6 million simulated value.
• Cost to scan a single smart contract: $122. Potential exploit revenue doubled every 1.3 months throughout 2025.
• AI agents operate autonomously: they reason through logic, write exploits, test against live blockchain state, and iterate – no human needed.

• $630 million drained across 27 exploits. Certik said this “could only be possible with AI.”
• The Kelp DAO hack ($292M) targeted bridge infrastructure, not Aave’s own code – a flaw standard audits miss. Consequence: $8.45 billion in withdrawals from Aave, TVL dropped 45%.

• Aave underwent 15+ audits, yet $14.5 billion still inside is a public target. Bigger protocols attract more AI compute power.
• Composability means you don't have to break Aave itself – just something it touches.

• OpenZeppelin distanced itself from Arios, viewing AI as a powerful defensive tool.
• Losses as a percentage of total value locked are declining: 1.49% projected for 2026, down from 7.24% in 2022.
• 99% of audits are correct; AI has a false positive rate above 97%.

• AI-assisted auditing: Helps defenders but also attackers – net advantage attacker.
• Formal verification: Can only prove specified properties, not bridge or supply chain attacks.
• Bug bounties: 76% of losses trace to North Korean state actors – a state won't be incentivized by a $5M reward.
• Insurance: Nexus Mutual holds only $170M in reserves against an $83–100B market – coverage under 1%.

• The Kelp DAO hack was stopped by the Arbitrum Security Council freezing $71M – the exact opposite of trustless code.
• Over $4 billion moved to Chainlink’s hardened infrastructure. The market votes for safety nets, not decentralization.
• CDFI (Centralized DeFi) uses security councils, KYC, human circuit breakers. Coinbase’s Morpho lending product has already surpassed $2.3 billion in originations.
• Regulators are licensing controlling persons behind protocols – effectively mandating centralization.

1. Check where admin keys live: Immutable contracts have pure code risk; upgradeable protocols have human/governance risk.
2. Understand composability: You inherit the risk of every bridge, oracle, and collateral your protocol touches.
3. Watch attrition: Over 40 DeFi protocols shut down in 2026 – not from hacks, but from the cost and futility of staying secure.
4. Treat insurance as a rounding error: Size positions as if there is no backstop – for most of the market, there isn’t.
5. Watch the drift toward CDFI: Regulations accelerate centralization by default.

• Arios’ warning may be overblown – loss rates are improving. But AI has permanently tilted the battlefield toward attackers.
• The only current solution is abandoning trustlessness for human safety nets – that is structurally bearish for DeFi’s original promise.
• AI didn’t kill DeFi, but it may have ended the era where “trustless” ever meant “safe.”

Summary follows...

Overview

In this video, the speaker analyzes the current Bitcoin price action and compares it to the 2018 bear market. He notes that the market structure is repeating itself and expects a short-lived rally followed by a decline.

• Current Price: Bitcoin is trading between $64,000 and $65,000, nearly unchanged from the previous week.
• Historical Patterns: In July of midterm years (e.g., 2018, 2022), there was often a rally of 20–40%. So far, the increase is about 10%.
• Comparison with 2018: The market structure is similar: a low in February, a higher low in March/April, another low in June/July, then a rally to the "Bear Market Resistance Band."
• Expectation: The speaker suspects Bitcoin will be rejected at this resistance level and fall again in August/September, similar to 2018.
• Feeling vs. Reality: Although the cycle feels different (due to the apathy top instead of euphoria), the structure is identical. The main difference is the lower price level.
• Retail Interest: Retail participation is declining, visible in falling YouTube views. Many are disappointed by altcoin promises and rug pulls.
• Bear Market Duration: Historically, the resistance band was not sustainably broken until the year after the halving (e.g., January 2019, March 2015). The bottom may still be ahead.

The speaker expects a rejection at the resistance band and a decline in late summer. Volatility may return, but a sustainable uptrend is not expected until the year after the halving.

In this Q&A session, various market topics are discussed. Here are the key points:

• The creator asks for feedback on speed and complexity of his content.
• 87% of viewers find the current format perfect.

• STRC is a preferred stock of MicroStrategy that should trade at par value of $100.
• Currently it is trading far below (~$86-87).
• If STRC does not return to par, it threatens dilution for common shareholders through issuance of new shares.
• The STRC model is a positive flywheel: New STRC shares → Bitcoin purchases → rising Bitcoin price.
• Bottom line: Without STRC recovery, MicroStrategy is 'cooked'.

• Bitcoin mining is currently extremely unprofitable: cost per Bitcoin ~$91,000, price ~$64,000.
• AI compute generates $1,600–$4,000 revenue per megawatt hour, Bitcoin mining only $80–$150.
• Therefore, many miners are pivoting to AI inference.
• Elon Musk's empire (Tesla, SpaceX) is massively focused on AI inference (e.g., Optimus, Cybercab, Megapacks).
• SpaceX successfully completed a Starship test flight – 200 launches could generate $1 trillion in revenue.
• Bottom line: Bitcoin mining in space is currently not viable.

• EchoStar (SATS) serves as a proxy for SpaceX, as SpaceX is not publicly traded.
• There is an arbitrage model: Fair value of EchoStar ranges from $81 to $159 depending on scenario, currently at $88.
• Risks: Complex capital structure, subsidiary bankruptcies, structural decline of core business.
• Do not buy LEAPS now, as SpaceX insiders will unlock more shares until November – price could drop to $100.
• Better to wait and buy cheaper.

• Marvel is cheaper than Nvidia: The pair chart shows a swap zone at 0.80–0.85.
• Jensen Huang (Nvidia CEO) predicts Marvel as the next $1 trillion company (6x from current $170B).
• Marvel benefits from the connectivity bottleneck (chips, memory, connectivity).
• Recommendation: Swap into Marvel at 0.85, back to Nvidia at 1.45.

• Broad ETFs contain many losers and only a few real winners.
• For a long horizon (children), individual stocks from the I13 basket are better.
• Capital expenditures (Capex) of hyperscalers (Amazon, Google, Meta, Microsoft) flow directly to I13 companies.
• This trend continues at least until 2028 – follow the money.
• Active strategies like pair trading or covered calls are possible in tax-free accounts.

• At age 32, financially independent, with 2–6 hours daily, you have ideal conditions.
• Recommended tools: Lilo (layer model), DCIS (DCA on steroids), IDSS (gold standard with confluence), ATR model.
• Important: Find your rhythm – practice one method 1000 times (Bruce Lee).
• Keep a trading journal, identify wins/losses, stay disciplined and patient.

• Tesla calls (Jan 2027, $400): Break-even at $450. Roll down to safer strikes or buy more time, as the last 3 months before expiry lose value quickly.
• Maximum MicroStrategy exposure: If Bitcoin market turns bullish in 60 days, rotate max 20% of your Bitcoin portfolio into MSTR.

• Concentration on disruptive assets (AI, Bitcoin, Space) pays off with a long time horizon.
• Patience, discipline, and consistent practice are the keys to success.

Summary of the Analysis of the Supplement That Reverses Biological Age

The study investigated whether dietary supplements provide additional longevity benefits for individuals who have already optimized their lifestyle (diet, exercise, sleep).

• Research Team: Kamal Pabis, Yuan Gruba, and Dr. Brian Kennedy from the National University of Singapore.
• Publication: The study "Supplements and Drugs Are Associated with Biological Age in a Cohort of Exceptionally Healthy Individuals" was published in 2026 in the journal Aging Cell.
• Participants: 4,260 self-proclaimed health enthusiasts who had purchased at least one epigenetic test.
• Key Characteristics: 71% of participants regularly took supplements – a much higher rate than the general population. The group also had very low smoking rates and a low incidence of chronic disease.
• Method: A cross-sectional observational study that measured the real-world association between supplement use and the biological clock (epigenetic clock).

• Biological Clock: A proprietary, saliva-based DNA methylation clock (9 CpG sites, accuracy: ± 5.4 years).
• Key Metric: The "Age Residual" – a negative value means a person is biologically younger than their chronological age.
• Most Effective Supplement: A commercial, delayed-release formulation of calcium alpha-ketoglutarate (AKG) with vitamins, known as Rejuvenate. It was associated with an average 1.8 years younger biological age.
• Critical Detail: Standard, non-delayed-release AKG showed no significant effect. The release mechanism is crucial for efficacy.

• Mechanism: AKG is a natural metabolite that declines with age. It is an essential substrate for TET enzymes, which act as the cell's cleanup crew, removing harmful methyl groups (epigenetic drift).
• Comparison to Other Supplements: NAD boosters (720 users) and Vitamin D3 (690 users) showed no significant correlation with a younger biological age. AKG, with only 143 users, showed a strong and significant effect.
• Coenzyme Q10 (CoQ10): Showed a trend, but lost statistical significance after final multivariate corrections.
• Supplement Stacking (Polypharmacy): Taking multiple supplements simultaneously showed no negative effects on biological age, challenging the idea that large stacks are inherently harmful.

• Conflict of Interest: Dr. Brian Kennedy, the senior author, is the Chief Scientific Officer of Ponce de Leon Health, the manufacturer of Rejuven AKG. However, the raw data shows the product has a robust effect despite a small sample size.
• Study Limitation: This is an observational study, which proves correlation, not causation. Randomized controlled trials (RCTs) are the gold standard, but are rarely funded for non-patentable supplements.
• Safety: AKG is generally well-tolerated. High doses may interact with medications. Individuals with kidney conditions or metabolic disorders should exercise caution.
• Dietary Sources: AKG is found in trace amounts in food, but not in concentrations that are therapeutically relevant. Supplementation is therefore practical.

• The study provides strong evidence that delayed-release calcium alpha-ketoglutarate is one of the most effective supplements for reducing biological age.
• The question remains whether this effect is replicable with other formulations or dosages. Discuss which dosage you use and if you have tracked any changes.

Bottom Line: Even with an optimized lifestyle, a specific supplement (delayed-release AKG) can offer a statistically significant and practically relevant benefit for biological youthfulness. Formulation and quality are paramount.

Summary: The Case of Tiffany Crawford

Late on the evening of May 2, 2017, Jason Crawford found his wife Tiffany shot dead in her car in front of their home in Cullman, Alabama. Initially deemed a suicide, forensic evidence soon raised serious doubts.

• Tiffany Crawford (née Webb, 1985) was a warm, caring woman who worked as a teacher and founded the Facebook group Mothers Helping Mothers.
• She was married to Jason Crawford, had two children and a stepson. The marriage was strained; Tiffany had an affair with Kyle Sunday and planned to leave Jason.
• In the weeks before her death, she seemed withdrawn, drank more alcohol, and appeared changed. Her affair ended when Kyle broke up with her.

• On the evening of May 2, Tiffany had dinner with a friend. Jason discovered messages about her affair on the computer and called her home.
• On the way, she called her mother, sounding tense but not panicked.
• A heated argument ensued outside the house. Jason refused to let her in, handed her work clothes, and closed the door. Shortly after, he heard two gunshots and found Tiffany dying in the car.

• Initial police response assumed suicide, but crucial steps were missed (no gunshot residue tests, no house search, the car was returned to Jason and cleaned).
• Jason's mother worked at the Cullman County Sheriff's Office, creating a conflict of interest. The case was transferred to the Alabama State Bureau of Investigation.
• Autopsy results (Dr. Valerie Green):
  • Two gunshot wounds to the head: one to the left jaw, one fatal to the left temple.
  • The first shot was not instantly fatal, but Tiffany would have been unable to fire a second shot.
  • No gunpowder residue on her skin, no contact wound.
  • The fatal shot was fired from at least 25 cm (10 inches) away – impossible in the cramped car if she had shot herself.
  • The gun was found in her left hand, though she was right-handed.
  • No blood on the window or door panel – the door must have been open.
  • Male DNA was found on the trigger and grip, but none of Tiffany's.

• Jason Crawford was charged with murder in May 2018; his trial began in November 2022.
• Prosecution: Marital problems, Tiffany's plans to leave, Jason's anger over the affair, his cold demeanor on the 911 call, and a failed polygraph test.
• Defense: Tiffany was suicidal (diary entry, therapy session), no blood on Jason, inconclusive DNA, and initial officers believed it was suicide.
• Jason's son Logan testified that he heard his father inside the house during the shots – the prosecution questioned his credibility.
• Jason admitted being angry and insulting Tiffany but denied killing her.
• The jury deliberated for less than 5 hours and found Jason guilty of intentional murder. He was sentenced to 99 years in prison.

The weight of evidence – especially forensic contradictions, lack of blood spatter, the shooting distance, and the gun in her left hand – strongly suggests murder. Tiffany's plans for the future and her state of mind make suicide improbable.

Opinion: The case is complex, but the evidence is convincing. The verdict appears just.

• Approximately 5,000 new tokens launch daily across major chains; on Pump.fun, an estimated 99% are purely value extraction.
• 85% of all tokens launched in 2025 were already trading below their launch price by February.
• Rug pulls stole between $2.8 and $5.8 billion in 2025 alone – and oversight is shrinking.

• Good risk: Solid team with a real product that might fail – that's normal investing.
• Rug: A product built specifically to take your money. This type is avoidable because the perpetrators leave traces.

1. Team & Partnerships

• Anonymous or fresh identities with no verifiable track record.
• AI-generated profile pictures, fabricated LinkedIn careers.
• Fake partnerships: A logo means nothing – verify through the partner's official channels.
• Example Scatman (July 2025): Fake SpaceX account, $135,000 stolen.

2. Lack of Real Work

• No working product, dead GitHub with no commits, whitepaper full of buzzwords.
• Real builders leave evidence: code, users, revenue, activity.

3. Tokenomics: Low Float, High FDV

• Low circulating supply (under 10-15% of total) creates artificial scarcity.
• Vesting cliffs: Huge token unlocks after 6-12 months – insiders have near-zero cost basis.
• Example Hawk Tuah: 96% of tokens held by insiders at launch, 89% crash.

4. On-Chain Checks

• Top 10 wallets holding more than 20% of supply → warning sign.
• Mantra (April 2025): 90% held by team/insiders, 90% drop in one day.
• Further points:
  • Liquidity locked for only a few days instead of months.
  • Mint function allows unlimited token creation.
  • Blacklist function (honeypot) prevents selling.
  • Fake renounced ownership: hidden admin powers.

5. Smart Contract Audits

• 75% of fraudulent projects have no audit; others use no-name firms or fake PDFs.
• Verify the audit on the auditor's domain, not on the project site. If the contract was redeployed after the audit, it's worthless.

6. Social Media & Community

• 70% of rugs use coordinated influencer campaigns: identical bullish posts at the same time.
• A legitimate community welcomes questions; a rug bans you for critical inquiries ("FUD").
• Unclear yield source in DeFi: If you don't know where the yield comes from, you are the yield.

1. Real Revenue: Protocols like GMX or Hyperliquid generate fees from genuine activity. Hyperliquid has crossed $1 billion in cumulative protocol revenue.
2. Sane Valuation: FDV under 15x annual revenue considered undervalued, above 50x expensive.
3. Smart Unlock Policy: Instead of insider dumping, revenue is used for buybacks (e.g., Hyperliquid).
4. Verifiable Users: Real daily active addresses, not ghost chains with inflated valuations.
5. Sustainable Emissions: Staking yields like Ethereum's 2.5-4%. Anything promising 400% without a clear source is a red flag.

1. Check Fundamentals: Does the protocol earn money from users or only from printed tokens? (DeFi Llama, Token Terminal)
2. Review Unlock Schedule: Look at next 30-90 days (e.g., token.unlocks.app). Reduce exposure 30 days before a cliff, reassess 14 days after.
3. Contract Scan: Use tools like RugCheck, Token Sniffer, Honeypot – hunt for mint, blacklist, hidden taxes.
4. Verify Liquidity: Don't trust badges – check actual lock duration on-chain (6-12 months or more desired).
5. Analyze Holder Concentration: Bubble Maps reveals connected wallets – one person pretending to be 40.

If a project trips three or more of these red flags (no audit, anonymous team, dodgy liquidity, etc.), walk away – no matter how good the narrative sounds. Less than 5% of rug pull losses are ever recovered. Prevention is the only strategy that works.