PJFP.com

Pursuit of Joy, Fulfillment, and Purpose

Author: PJFP

  • Tutorial: Removing Google Nano Banana (SynthID) Watermarks from AI-Generated Images Using Free Adobe Express Tools

    Before

    After

    Important Disclaimer: Watermarks like Google’s SynthID (embedded in images generated by Nano Banana Pro or Gemini’s image tools) exist to promote transparency and responsible use of AI-generated content. Removing them may violate Google’s terms of service, copyright laws, or platform policies — especially if the image isn’t yours or is used commercially without permission. This tutorial is for educational and personal fair-use purposes only. Always respect intellectual property rights and consider legitimate alternatives (e.g., Google’s paid Ultra plan for watermark-free exports). Proceed at your own risk.

    TL;DR

    You can remove the visible Gemini/Nano Banana watermark (the little sparkle/diamond logo) completely for free using Adobe Express’s crop or AI Remove/Spot Healing tool. The invisible SynthID watermark cannot be fully removed with free tools — only diluted slightly through editing/exporting. The whole process takes 5–15 minutes per image.

    Key Takeaways

    • Visible watermark → easily removed with cropping or Adobe Express free “Remove object” / Spot Healing tool
    • Invisible SynthID → not fully removable without paid/specialized tools; editing only reduces detection confidence a little
    • Adobe Express free tier works perfectly for this and lets you export without its own watermark if you avoid premium assets
    • Always keep the original file and disclose AI origin when sharing
    • Better long-term solution: pay for Gemini Ultra / Nano Banana Pro to get clean exports natively

    Detailed Step-by-Step Tutorial

    Step 1: Get Your Nano Banana Image

    1. Open Gemini (web or app) → Nano Banana
    2. Generate your image
    3. Download it (free tier includes visible watermark)

    Step 2: Open Free Adobe Express

    Go to adobe.com/express → Sign in with free Adobe account → “Start for free”

    Step 3: Quickest Method – Crop It Out

    1. Upload your image
    2. Use the Crop tool → drag to exclude the bottom-right corner watermark
    3. Apply → Done (perfect for most images)

    Step 4: Remove Visible Watermark with AI (When Cropping Isn’t Possible)

    1. In the left panel → Quick Actions → “Remove object” (or search “remove”)
    2. Brush over the Gemini sparkle logo
    3. AI automatically fills the area with surrounding pixels
    4. Repeat or use Clone Stamp if needed

    Step 5: Export Without Adobe Watermark

    1. Click Download
    2. Choose PNG or JPG
    3. If it tries to add Adobe watermark → you probably used a premium template/element → undo and use only free assets, or toggle watermark off in settings
    4. Free basic edits export clean 99% of the time

    Step 6: Verify

    • Zoom in → no visible logo
    • Optional: upload to Hive Moderation or ask Gemini “Is this AI-generated?” → invisible SynthID usually still detectable

    Alternative Free Tools if Adobe Express Is Acting Up

    • WatermarkRemover.io (4 free removals/day)
    • Photopea.com (web Photoshop clone)
    • Photoshop Express mobile app (free Spot Heal)
    • GIMP (desktop, fully free)

    My Thoughts on This Whole Thing

    Google adding both visible and invisible watermarks is actually a good move for transparency — the problem is they lock clean exports behind the priciest Ultra tier. For hobbyists and educators who just want to use a nice AI image in a presentation or blog without an ugly logo in the corner, having to pay $20+/month feels excessive.

    Adobe Express giving us a powerful, free “Remove object” tool essentially hands everyone a workaround for the visible mark, which is why this method works so well right now. The invisible SynthID is much harder to defeat without specialized (often paid or legally gray) tools, so for most practical purposes, the images are still identifiable as AI-generated — which keeps the transparency promise somewhat intact.

    Ethically, I’m fine with individuals cleaning up images they generated themselves for personal or clearly disclosed use. The line gets crossed when people start stripping watermarks to pass off AI art as human-made photography or to sell commercially without disclosure.

    Long-term, I hope Google adds a middle-tier plan that removes the visible logo (keep SynthID) for a few bucks a month — that would solve 95% of the frustration without undermining their transparency goals.

    Until then… crop tool go brrr. 🐒

  • The King of Hollywood: 7 Lessons on Power and Persuasion from Michael Ovitz and David Senra

    When the co-founder of Creative Artists Agency (CAA) sits down with David Senra, the host of the Founders podcast, you don’t just get industry gossip—you get a masterclass in agency, psychology, and relentless ambition. Michael Ovitz, often cited as the most powerful man in Hollywood during the 1980s and 90s, shared the playbook he used to revolutionize the entertainment industry.

    From his early days in the mailroom to orchestrating the sale of Columbia Pictures to Sony, Ovitz’s career is a testament to the power of information and relationships. Below is a breakdown of his conversation with David Senra, including key takeaways and a detailed summary of their discussion.

    TL;DW

    Michael Ovitz argues that success is driven by “frame of reference”—the accumulation of experiences that allows you to instinctively spot quality and talent. He emphasizes that fear is the enemy of business, that you must relentlessly study history to leverage it in the present, and that true salesmanship often involves “punching without punching”—selling without ever explicitly asking for the sale.

    Key Takeaways

    • Build a “Frame of Reference”: You cannot spot excellence if you haven’t seen it before. Ovitz believes in consuming vast amounts of information—art, culture, business history—to build a mental database that allows for instant pattern recognition.
    • Information is Leverage: As a mailroom trainee, Ovitz showed up at 6:30 AM (hours before anyone else) to read the agency’s private files. This gave him an encyclopedic knowledge of the business that his peers lacked.
    • The “No Guardrails” Mindset: Creativity in business means refusing to accept arbitrary boundaries. As Ovitz famously states, “I’ve never seen a guardrail I don’t try to jump”.
    • Punching Without Punching: The highest form of sales is demonstrated by David Rockefeller, who raised millions for MoMA without ever asking Ovitz for a dime. He simply built a relationship and shared a vision until Ovitz wanted to contribute.
    • Radical Transparency creates Loyalty: At CAA, Ovitz instituted a rule of “no lying.” If an agent didn’t know an answer, they had to say “I don’t know” and follow up later. This created trust in an industry famous for dishonesty.

    Detailed Summary

    1. The Mailroom Strategy: Outworking the Competition

    Ovitz’s career began in the mailroom at William Morris. Realizing he had no nepotistic connections in a relationship-driven town, he decided to differentiate himself through pure knowledge. While the other trainees arrived at 9:00 AM, Ovitz arrived at 6:30 AM.

    He read the correspondence of the top agents, learning the history of the industry. This allowed him to speak the language of the older generation of filmmakers. When he later met legendary directors, he could discuss their obscure influences (like Frank Capra or Howard Hawks) because he had done the reading. He noted that he wasn’t necessarily smarter than the Ivy League trainees, but he eradicated them by outworking them.

    2. The “Frame of Reference”

    A recurring theme in the interview is the “frame of reference.” Ovitz explains that his ability to spot talent—whether it was a young Wolfgang Puck in a parking lot restaurant or the chef Nobu Matsuhisa—came from constantly scanning the world for excellence.

    He creates a “personal AI” in his brain by consuming hundreds of images of art, reading widely, and meeting people. This creates a benchmark. When he met Nobu, he knew the chef was special not just because the food was good, but because Nobu “filled the room” with a sensei-like presence.

    3. The Coca-Cola Deal and The $3 Million Check

    One of the most tactical examples of Ovitz’s negotiation style involved Coca-Cola. CAA took over Coke’s advertising, employing film directors to make commercials—a move the industry mocked. When Coke sent CAA a check for $3 million to cover the cost of a specific commercial, Ovitz sent it back voided.

    He told them the commercial only cost $30,000 (having been made on an Apple IIe computer). He refused to let the client overpay for the production, which established immense trust. He then told them, “You’re not going to overpay for commercials, but you got to pay us.” This move allowed him to negotiate a much higher fee for the agency’s intellectual property and strategy rather than just production margins.

    4. Lessons from Mentors: Rockefeller and Morita

    Ovitz collected mentors as aggressively as he collected art. Two stand out:

    • David Rockefeller: Ovitz learned the art of the “soft sell.” Rockefeller invited Ovitz to join the MoMA board and spent hours discussing art and architecture, never bringing up money. By the end, Ovitz wrote a larger check than he ever intended, purely out of respect for Rockefeller’s integrity and vision.
    • Akio Morita (Sony): Ovitz admired Morita’s courage to disrupt his own business. Morita taught him the value of “thinking big”—not just building a company, but changing the perception of a nation (Japan). Ovitz also recounted how Morita hired his harshest critic, Norio Ohga, because he valued an honest “mirror” over a “yes man”.

    5. The Friendship with Michael Crichton

    Ovitz speaks touchingly of his 30-year friendship with author Michael Crichton. He describes Crichton as possessing a unique work ethic: he wouldn’t write every day, but when a deadline approached, he would write 20 hours a day for months. Crichton wrote Jurassic Park in a five-month burst of intensity. The biggest lesson Ovitz took from Crichton was “curiosity about everything”.

    Some Thoughts

    What stands out most in this interview is the bridge Ovitz builds between the “old world” of Hollywood and the “new world” of Silicon Valley. He speaks about Marc Andreessen and Ben Horowitz with the same reverence he holds for Paul Newman or Martin Scorsese.

    Ovitz’s philosophy is ultimately one of input/output. He treats his brain like a machine learning model—if you feed it high-quality data (art, history, business biographies), it will output high-quality decisions (spotting Nobu, packaging Jurassic Park). In an age of algorithmic curation, Ovitz represents the value of manual curation—going to the library, reading the files, and seeing the world with your own eyes.

    As he told Senra regarding his relentless drive even after achieving wealth: “I’ve never seen a guardrail I don’t try to jump”. For entrepreneurs, that is the only way to operate.

  • All-In Podcast Recap: Epstein Files, Tether’s Billions, Nvidia Accounting & Poker Psychology

    Live from The Venetian: The Besties break down the Epstein file release, the massive margins of Tether, the Michael Burry vs. Nvidia debate, and a masterclass in risk with Alan Keating.

    In this special live episode recorded during the F1 weekend in Las Vegas, the “Besties” (Chamath Palihapitiya, Jason Calacanis, David Sacks, and David Friedberg) reunite in person. The agenda is packed: political intrigue surrounding Jeffrey Epstein, the financial dominance of stablecoins, technical debates on AI chip accounting, and high-stakes poker strategy.

    TL;DR: Executive Summary

    The US government has voted nearly unanimously to release the Epstein files, leading the hosts to speculate that the lack of leaks points to intelligence agency involvement rather than political dirt on Donald Trump. Chamath details a meeting with Tether CEO Paolo Ardoino, revealing a business holding over $100 billion in US Treasuries with profit margins potentially exceeding 95%. The group then debates Michael Burry’s short position on Nvidia, with Friedberg defending the “useful life” of AI chips under GAAP accounting. Finally, poker legend Alan Keating joins to discuss “soul reading” opponents and mastering fear in high-stakes games.

    Key Takeaways

    • The Epstein Intelligence Theory: The hosts argue that if the files contained damaging information on Donald Trump, it would have been leaked during the Biden administration. The prevailing theory discussed is that Epstein may have been an intelligence asset (CIA/Mossad/Russia), explaining the long-standing secrecy.
    • Tether is a Financial Juggernaut: Tether holds approximately $135 billion in US Treasuries and operates with roughly 100 employees. Chamath estimates the business runs at 95%+ margins, effectively exporting US dollar stability to developing nations while capturing massive interest yields.
    • Nvidia vs. Michael Burry: “The Big Short” investor Michael Burry is shorting the sector, arguing tech companies are “cooking the books” by depreciating AI chips over 6 years when they become obsolete in 3. Friedberg counters that chips retain a “useful life” for inference and background tasks long after they are no longer top-of-the-line.
    • Google Gemini 3: Google has regained the lead on LLM benchmarks with Gemini 3. The conversation highlights a shift toward proprietary silicon (TPUs) and a fragmented chip market, posing a potential long-term risk to Nvidia’s dominance.
    • The “Oppenheimer” Moment: David Friedberg reveals he decided to return as CEO of Oho after watching the movie Oppenheimer, realizing he needed to be an active operator rather than a passive board member.

    Detailed Episode Breakdown

    1. The Epstein Files Release

    In a stunning bipartisan move, the House and Senate voted nearly unanimously to release the Epstein files. The Besties analyzed why this is happening now. Sacks and Chamath suggested that because Epstein was the “most investigated human on earth,” any compromising information regarding Trump would likely have been weaponized politically by now.

    The discussion pivoted to the source of Epstein’s wealth. Chamath noted Epstein managed money for billionaires and charged inexplicable fees for “tax advice”—such as a documented $168 million payment from Apollo’s Leon Black. The hosts speculated that Epstein likely functioned as a spy or asset for intelligence agencies, which would explain the protective layer surrounding the files for so long.

    2. Tether and the Stablecoin Boom

    Chamath shared insights from a dinner with Tether CEO Paolo Ardoino. Tether’s financials are staggering: approximately $135 billion in US Treasuries and billions more in Bitcoin and gold.

    The hosts discussed the utility of stablecoins in high-inflation economies, where locals use USDT to preserve purchasing power. Because Tether earns the interest on the backing treasuries (rather than passing it to the coin holder), and operates with a lean team, the company generates billions in pure profit. Sacks noted that future US regulations might eventually force stablecoin issuers to share that yield with users, but for now, it remains one of the most profitable business models in the world.

    3. Accounting Corner: Is Nvidia Overvalued?

    Michael Burry is shorting the semiconductor sector, claiming companies are inflating earnings by depreciating Nvidia chips over 6 years despite rapid technological obsolescence.

    Friedberg launched a segment dubbed “Accounting Corner” to rebut this. He explained that under GAAP standards, an asset’s useful life is determined by its ability to generate revenue, not just its technological superiority. Even if an H100 chip isn’t the fastest on the market in year 4, it can still run inference models or handle lower-priority compute tasks, justifying the longer depreciation schedule. Chamath added that tech giants monitor “output tokens” closely; if a chip wasn’t profitable, they would simply turn it off.

    4. Poker Strategy with Alan Keating

    The episode concluded with Alan Keating, a high-stakes poker player famous for his loose, aggressive style. Keating explained his philosophy, which relies less on “solvers” (GTO strategy) and more on “soul reading”—navigating the fear and psychology of the table.

    He broke down a famous hand where he beat Doug Polk with a 4-2 offsuit, explaining that he sensed fear in Polk’s betting patterns on the turn. Keating described his approach as finding “beauty in the chaos” and dragging opponents into “deep water” where they are uncomfortable and prone to errors.

    Editorial Thoughts

    This episode marked a distinct shift in the podcast’s tone regarding crypto, moving from general skepticism to a recognition of the sheer scale and utility of stablecoins like Tether. The “Accounting Corner” segment, while technical, provided critical context for investors trying to value the AI stack—suggesting the AI boom has more fundamental accounting support than bears like Burry believe. Finally, the live format from Las Vegas brought a looser, more energetic dynamic to the conversation, highlighting the chemistry that makes the show work.

  • Robinhood CEO Vlad Tenev on “Vibe Trading,” Prediction Markets, and Democratizing Private Equity

    In a recent discussion on the Uncapped podcast with Jack Altman, Robinhood co-founder and CEO Vlad Tenev opened up about the company’s transition from a trading platform to a “financial super app.” Tenev discussed the explosion of prediction markets, the role of AI in creating “vibe trading,” and his vision for tokenizing private assets to help retail investors capture value earlier.

    TL;DR

    Robinhood is aggressively expanding beyond simple stock trading. Vlad Tenev highlights three major frontiers: the rise of prediction markets as “truth machines,” the use of AI to create autonomous “vibe trading” experiences, and the tokenization of private assets to allow everyday investors access to companies like SpaceX or OpenAI before they go public.

    Key Takeaways

    • From App to Ecosystem: Robinhood no longer views itself merely as a trading platform but as a “financial home” and super app, encompassing banking, credit cards, and retirement accounts.
    • Prediction Markets are Booming: Tenev views prediction markets not just as speculation, but as “truth machines” that offer cleaner data than traditional polling or media. Robinhood’s volume in this sector has seen massive growth.
    • “Vibe Trading”: Tenev coined the term “vibe trading” to describe a future where AI agents manage a user’s portfolio based on high-level intent, risk tolerance, and personal goals rather than manual trade execution.
    • Solving the Private Equity Gap: Tenev argues that the biggest inequity in modern markets is that value now accrues in private markets (e.g., SpaceX, OpenAI) rather than public ones. He believes tokenization is the solution to give retail investors access.
    • Generational Shifts: Contrary to stereotypes, Gen Z is opening retirement accounts as early as 19 years old, signaling a shift toward financial conservatism compared to millennials.

    Detailed Summary

    The Evolution of the Brokerage

    Tenev traces the history of the online brokerage from the deregulation of commissions in 1975 (the “Mayday” event that birthed Charles Schwab) to the mobile-first revolution led by Robinhood. While early digital brokers like E-Trade catered to Gen X, Robinhood capitalized on two shifts: the ubiquity of mobile phones and the infrastructure changes brought by high-frequency trading, which lowered costs enough to offer commission-free trading.

    Today, Robinhood generates over a billion dollars in revenue across multiple business lines, aiming to be the primary financial institution for its users.

    Prediction Markets: The “Truth Machines”

    One of the fastest-growing segments for the company is prediction markets. Tenev notes that the 2024 Presidential Election was a “Big Bang” moment for the industry, validating these markets as superior forecasting tools compared to traditional polls.

    He argues that because participants have “skin in the game,” prediction markets filter out noise and bias, acting as “truth machines.” Beyond politics, this is expanding into sports and entertainment, which Tenev views as an inevitability in an economy where AI automates traditional labor.

    Tokenization and Private Markets

    Tenev expressed deep concern regarding where economic value is created today versus thirty years ago. When Microsoft and Apple went public, they were valued in the low billions, allowing public market investors to capture the majority of their growth. Today, companies like SpaceX or OpenAI may reach trillion-dollar valuations while still private, shutting out retail investors.

    His solution is tokenization. Similar to how stablecoins operate, Tenev envisions a structure where private securities are held in a “bucket” while tokens representing them trade freely 24/7 on a blockchain. This would democratize access to private equity, a move he sees as the eventual end-state of capital markets.

    AI and the Era of “Vibe Trading”

    Robinhood is heavily integrating AI into its operations, achieving high deflection rates in customer support and increased coding output from engineering. However, the consumer-facing future is what Tenev calls “Vibe Trading.”

    In this model, the user interface shifts from manual execution to intent-based directives. A user might tell an AI agent their risk appetite, long-term goals, and interests, and the agent—acting as a “financial home”—executes the strategy. Tenev believes this will also solve mundane friction points, such as AI agents automatically handling the paperwork to switch bank accounts.

    Thoughts on the Interview

    Vlad Tenev’s commentary suggests a significant pivot in Robinhood’s brand identity. Originally seen as the disruptor that “gamified” trading, the company is now positioning itself as the mature “financial super app” for a generation that is aging into wealth.

    The most compelling insight is the focus on tokenization. Tenev correctly identifies that the “public market” is no longer the primary engine of wealth creation for early-stage innovative companies. If Robinhood can successfully navigate the regulatory hurdles to tokenize private equity (essentially breaking down the walls of the accredited investor requirements via technology), they wouldn’t just be a brokerage; they would fundamentally alter the structure of modern capitalism.

    Furthermore, the concept of “Vibe Trading” aligns with the broader tech trend of “agentic AI.” It moves the user value proposition from “we give you the tools to do it yourself” to “we have the intelligence to do it for you,” which may appeal to a broader demographic than active traders.

  • NVIDIA (NVDA) Q3 FY2026 Earnings: $57B Record Revenue, Blackwell “Off the Charts,” $65B Guidance – The AI Boom Is Still Accelerating

    November 20, 2025 – NVIDIA just delivered the most dominant quarter in the history of tech and told the world the next one will be even bigger. The market is partying like it’s 2021.

    TL;DR

    • Revenue $57.01B (+62% YoY, beat by ~$1.8–2B)
    • Data Center $51.2B (+66% YoY, +$10B sequentially) – now 90% of total revenue
    • GAAP EPS $1.30 (+67% YoY)
    • Q4 guidance $65B (±2%) – obliterates street $61.98B (some buyside whispers were $75B → Jensen sandbagging again)
    • Blackwell sales “off the charts”, cloud GPUs completely sold out for the foreseeable future
    • CFO Colette Kress confirmed ≈$500B Blackwell + Rubin revenue visibility 2025–2026 (analysts now calling it $500B pipeline through FY2027)
    • Gross margin 73.6% (tiny miss due to Blackwell ramp costs), guided back to 75.0% next quarter
    • Free cash flow $22.1B in a single quarter
    • Top 4 customers = 61% of revenue (22% / 15% / 13% / 11%) – concentration risk is real but demand makes it a feature
    • Stock ripped +5.5% after-hours → +$220B+ market cap in minutes, lifting entire AI complex

    Key Takeaways

    • Demand is not slowing — it’s compounding. Jensen: “Compute demand keeps accelerating and compounding across training and inference — each growing exponentially. We’ve entered the virtuous cycle of AI.”
    • Blackwell ramp is unprecedented – already the majority of new Data Center mix, sold out for months, driving the entire $10B sequential jump
    • Gaming ($4.3B) and Automotive ($592M) missed estimates → literally nobody cares when Data Center grew $10B in one quarter
    • Customer concentration: Four hyperscalers = 61% of revenue. Everyone knows who they are. Everyone also knows they can’t build without NVIDIA
    • Margins dipped to 73.6% only because of Blackwell complexity/HBM costs – guided 75% next quarter, street relieved
    • Balance sheet is absurd: $60.6B cash + $22.1B quarterly FCF. Berkshire is only ~$320B ahead
    • Physical AI multi-trillion opportunity already “multi-billion” today

    Detailed Summary

    NVIDIA printed $57.01 billion in a single quarter — a number larger than the entire annual revenue of 99% of public companies. Data Center alone did $51.2 billion (+66% YoY, +$10 billion sequentially). Let that sink in.

    Blackwell is not “ramping” — it’s exploding. It is already the majority of new Data Center revenue and cloud providers are in a literal bidding war for every wafer. Jensen was blunt: “Blackwell sales are off the charts, and cloud GPUs are sold out.”

    Yes, Gaming and Automotive missed estimates (who cares), Pro Visualization crushed it (+56% YoY), but the only number that matters is the $500 billion in confirmed Blackwell + Rubin orders the company can already see through calendar 2026 (Bloomberg Intelligence now calling it $500B pipeline through fiscal 2027).

    China export restrictions? Effectively $0 impact in guidance. The rest of the planet is making up for it and then some — sovereign AI factories, enterprises, everyone is building.

    Networking (Spectrum-X + InfiniBand) up ~162% YoY to $8.2B+ — the hidden monster line item nobody talks about.

    Market & Analyst Reaction

    Initial spike was +4%, then kept climbing → closed extended trading up ~5.5%, adding north of $220 billion in market cap. Entire AI food chain ripping: CoreWeave +4%, Nebius +4%, AMD +2%, Micron +2%, Broadcom +2%, Super Micro +8%.

    Goldman Sachs (James Schneider) first note post-earnings:

    “Strong quarter with upside to guidance should provide relief for the stock… We expect the stock to trade higher following a stronger quarter and guidance relative to the Street.”

    X was pure euphoria last night – here are some of the top posts (all >5K likes):

    • https://x.com/EconomyApp/status/1991259207878996127 ← Clean chart
    • https://x.com/Quartr_App/status/1991259508941734389 ← Jensen quote card
    • https://x.com/KobeissiLetter/status/1991255966235419112 ← +$205B market cap meme
    • https://x.com/amitisinvesting/status/1991263435493974047 ← Full breakdown thread
    • https://x.com/FromValue/status/1991275128439123451 ← “NVIDIA just printed more FCF than most companies make in revenue”

    My Thoughts

    This was a “relief rally on steroids”. Anyone still waiting for the AI capex slowdown just got obliterated. The $500 billion visibility isn’t hopium — it’s what they can already see in purchase orders.

    The moat is now impenetrable: CUDA + NVLink + Spectrum-X + Grace CPU + Blackwell/Rubin roadmap = Microsoft Windows-level lock-in for the AI era.

    At ~44× forward earnings the stock looks expensive until you realize the base case is now ~$260–280B annual revenue run-rate by late 2026. That puts the multiple in the low 20s. That is no longer the bull case — that’s the new floor.

    The Christmas rally is officially back on. NVIDIA just saved it.

  • Pershing Square’s Bold Plan: Relist Fannie Mae & Freddie Mac on NYSE in November 2025 – Taxpayers Could Gain $300B+

    Pershing Square’s Bold Plan: Relist Fannie Mae & Freddie Mac on NYSE in November 2025 – Taxpayers Could Gain $300B+

    TL;DR:

    Bill Ackman’s Pershing Square Capital Management just released a 28-page investor presentation urging the Trump administration to immediately (1) deem the Treasury’s Senior Preferred Stock repaid, (2) exercise the 79.9% warrants, and (3) relist Fannie Mae (FNMA) and Freddie Mac (FMCC) on the NYSE — all while keeping the GSEs in conservatorship. They claim this can be done before the end of November 2025 and would instantly value the U.S. taxpayer’s stake at over $300 billion without disrupting mortgage affordability.

    Key Takeaways

    • Fannie & Freddie OTC shares have already more than doubled in 2025 on Trump administration statements.
    • The three-step plan (repay SPS → exercise warrants → NYSE relisting) can be executed immediately by Treasury and FHFA.
    • Post-relisting, Treasury would own 79.9% of two NYSE-listed companies worth a combined ~$387 billion (Pershing estimate).
    • Taxpayers have already received $301 billion in dividends — $25 billion more than required under the original 10% deal.
    • Pershing strongly opposes any conversion of Senior Preferred into common — calls it value-destructive and legally risky.
    • Relisting unlocks massive institutional buying (many funds are barred from OTC stocks) and fulfills Trump’s campaign promise timing.
    • Conservatorship continues for years, giving the administration runway to finalize capital rules, backstop structure, and governance.

    Detailed Summary of the Pershing Square Presentation (November 2025)

    In a presentation titled “Promises Made, Promises Kept”, Pershing Square lays out a politically and financially attractive path for the second Trump administration to deliver on its GSE reform pledges without raising mortgage rates or rushing a full privatization.

    The core argument: the government has already been fully repaid (and then some) via $301 billion of dividends since 2008. The Obama-era 2012 “Net Worth Sweep” was paused under Mnuchin, but never fully reversed. Pershing says a simple letter agreement between Treasury and FHFA can officially retire the Senior Preferred Stock today.

    Once the SPS is gone, Treasury can exercise its long-held warrants for 79.9% of the common stock at essentially zero cost. The GSEs already meet every NYSE listing requirement (market cap, float, share price, shareholder count, etc.). FHFA can approve relisting while keeping full conservatorship powers intact — no change to operations, no new capital raises, no dividend payments to juniors until fully recapitalized.

    Pershing’s valuation math (as of 12/31/2025):

    • Fannie Mae: 16× 2026E EPS → ~$42–45/share → Treasury 79.9% stake ≈ $196 billion
    • Freddie Mac: 13× 2026E EPS → ~$44/share → Treasury 79.9% stake ≈ $114 billion
    • Total taxpayer value: >$310 billion (plus junior preferred)

    They explicitly reject the idea of converting Senior Preferred into common, warning it would trigger new litigation, force government consolidation onto the federal balance sheet, and slash valuations by 27–56% depending on the multiple the market would assign to a company that wiped out private shareholders.

    My Thoughts

    This is classic Ackman: aggressive, detailed, and perfectly timed to influence policy while he has a massive economic interest (Pershing owns large common positions in both GSEs). The beauty of the proposal is that it is genuinely low-risk from a mortgage-market standpoint and gives the administration an instant “win” before Thanksgiving 2025.

    The politics line up perfectly: Trump gets to post on Truth Social that he turned two “bailed-out” companies into a $300 billion+ taxpayer windfall, keeps 30-year mortgage rates stable (or even lower), and still retains total control to shape the final exit over the next three years.

    If Treasury and FHFA actually follow the three steps before November 30, 2025, the OTC-to-NYSE pop could be one of the largest wealth-transfer events in market history — and almost entirely to existing common shareholders (retail + hedge funds that held on since 2008).

    Watch for any joint Treasury/FHFA announcement or letter agreement in the next two weeks. That will be the trigger.

    Disclosure: Like Pershing Square, the author may have direct or indirect exposure to FNMA/FMCC securities.

  • Google Launches Gemini 3 Pro (Nov 18, 2025): The Most Powerful Agentic & Reasoning Model Yet – Now Available for Developers

    TL;DR


    Google just released Gemini 3 Pro – their smartest model ever. It crushes benchmarks in reasoning, coding, agentic workflows, and multimodal understanding. New tools include Google Antigravity (free agentic IDE), better bash/tool-calling, 1M context, and “vibe coding” that turns a single natural-language prompt or sketch into a full working app. Available today in Google AI Studio (free with limits) and via Gemini API at $2/$12 per million tokens.

    Key Takeaways

    • Gemini 3 Pro is Google’s new flagship model (November 18, 2025) with state-of-the-art reasoning and agentic capabilities
    • Tops almost every major benchmark, including #1 on WebDev Arena (1487 Elo) and 54.2% on Terminal-Bench 2.0
    • New Google Antigravity – free public preview agentic development platform for Mac/Windows/Linux
    • 1 million token context window + significantly better long-context usage than Gemini 2.5 Pro
    • Best-in-class multimodal: new SOTA on MMMU-Pro (image) and Video MMMU
    • Advanced “vibe coding”: build entire interactive apps/games from one prompt, voice note, or napkin sketch
    • New client-side & server-side bash tools, structured outputs + grounding, granular vision resolution control
    • Pricing (preview): $2/M input tokens, $12/M output tokens (≤200k context), free tiered after that
    • Free access (rate-limited) inside Google AI Studio right now
    • Already integrated into Cursor, Cline, JetBrains, Android Studio, GitHub, Emergent, OpusClip and many more

    Detailed Summary of the Gemini 3 Launch

    On November 18, 2025, Google officially introduced Gemini 3 Pro, calling it their “most intelligent model” to date. Built from the ground up for advanced reasoning and agentic behavior, it outperforms every previous Gemini version and sets new records across coding, multimodal, and general intelligence benchmarks.

    Agentic Coding & Google Antigravity

    The biggest highlight is the leap in agentic coding. Gemini 3 Pro scores 54.2% on Terminal-Bench 2.0 (vs 32.6% for Gemini 2.5 Pro) and handles complex, long-horizon tasks across entire codebases with far better context retention.

    To showcase this, Google launched Google Antigravity – a brand-new, completely free agentic development platform (public preview for macOS, Windows, Linux). Developers act as architects while multiple autonomous agents work in parallel across editor, terminal, and browser, producing detailed artifacts and reports.

    Vibe Coding & One-Prompt Apps

    Gemini 3 Pro finally makes “vibe coding” real: describe an idea in plain English (or upload a sketch/voice note) and get a fully functional, interactive app in seconds. It currently sits at #1 on WebDev Arena with 1487 Elo. Google AI Studio’s new “Build mode” + “I’m feeling lucky” button lets anyone generate production-ready apps with almost zero code.

    Multimodal Leadership

    • New SOTA on MMMU-Pro (complex image reasoning) and Video MMMU
    • Advanced document understanding far beyond OCR
    • Spatial reasoning for robotics, XR, autonomous vehicles
    • Screen understanding + mouse-movement intent detection (Visual Computer demo)
    • High-frame-rate video reasoning

    Gemini API & Developer Tools Updates

    • New client-side and hosted server-side bash tools for local/system automation
    • Grounding + URL context can now be combined with structured outputs
    • Granular control over vision fidelity (trade quality vs latency/cost)
    • New “thinking level” parameter and stricter thought-signature validation for reliable multi-turn reasoning

    Pricing & Availability (as of Nov 18, 2025)

    • Gemini API (Google AI Studio & Vertex AI): $2 per million input tokens, $12 per million output tokens (prompts ≤200k tokens)
    • Free tier with rate limits in Google AI Studio
    • Immediate integration in Cursor, Cline, JetBrains, Android Studio, GitHub Copilot ecosystem, Emergent, OpusClip, etc.

    My Thoughts

    Gemini 3 Pro feels like the moment AI coding agents finally cross from “helpful assistant” to “can run an entire sprint by itself.” The combination of 1M context, 54% Terminal-Bench, and the new Antigravity IDE means developers can now delegate whole features or refactors to agents and actually trust the output.

    The “vibe coding” demos (retro game from one prompt, full app from a hand-drawn sketch) are no longer parlor tricks – they are production-ready in Google AI Studio today. For indie hackers and prototyping teams this is an absolute game-changer.

    Google pricing remains extremely aggressive ($2/$12) compared to some competitors, and giving Antigravity away for free is a bold move that will pull a huge portion of the agentic-dev-tool market toward their ecosystem overnight.

    If you develop, design, or just have ideas – go download Antigravity and play with Gemini 3 Pro in AI Studio right now. 2026 is going to be built with this model.

    Get started:
    → Google AI Studio (free)
    → Google Antigravity download

  • Satya Nadella on AI Adoption, Agentic Commerce, and Why This CapEx Boom Is Different From the Dot-Com Bubble (Cheeky Pint Interview Nov 2025)


    Microsoft CEO Satya Nadella sat down with Stripe co-founder John Collison on the Cheeky Pint podcast in November 2025 for a wide-ranging, candid conversation about enterprise AI diffusion, data sovereignty, the durability of Excel, agentic commerce, and why today’s AI infrastructure build-out is fundamentally different from the 2000 dot-com bust.

    TL;DW – The 2-Minute Version

    • AI is finally delivering “information at your fingertips” inside enterprises via Copilot + the Microsoft Graph
    • This CapEx cycle is supply-constrained, not demand-constrained – unlike the dark fiber of the dot-com era
    • Excel remains unbeatable because it is the world’s most approachable programming environment
    • Future of commerce = “agentic commerce” – Stripe + Microsoft are building the rails together
    • Company sovereignty in the AI age = your own continually-learning foundation model + memory + tools + entitlements
    • Satya “wanders the virtual corridors” of Teams channels instead of physical offices
    • Microsoft is deliberately open and modular again – echoing its 1980s DNA

    Key Takeaways

    • Enterprise AI adoption is the fastest Microsoft has ever seen, but still early – most companies haven’t connected their full data graph yet
    • Data plumbing is finally happening because LLMs can make sense of messy, unstructured reality (not rigid schemas)
    • The killer app is “Deep Research inside the corporation” – Copilot on your full Microsoft 365 + ERP graph
    • We are in a supply-constrained GPU/power/shell boom, not a utilization bubble
    • Future UI = IDE-style “mission control” for thousands of agents (macro delegation + micro steering)
    • Agentic commerce will dominate discovery and directed search; only recurring staples remain untouched
    • Consumers will be loyal to AI brands/ensembles, not raw model IDs – defaults and trust matter hugely
    • Microsoft’s stack: Token Factory (Azure infra) → Agent Factory (Copilot Studio) → Systems of Intelligence (M365 Copilot, GitHub Copilot, Security Copilot, etc.)
    • Culture lesson: don’t let external memes (e.g. the “guns pointing inward” cartoon) define internal reality

    Detailed Summary

    The conversation opens with Nadella’s excitement for Microsoft Ignite 2025: the focus is no longer showing off someone else’s AI demo, but helping every enterprise build its own “AI factory.” The biggest bottleneck remains organizing the data layer so intelligence can actually be applied.

    Copilot’s true power comes from grounding on the Microsoft Graph (email, docs, meetings, relationships) – something most companies still under-utilize. Retrieval, governance, and thick connectors to ERP systems are finally making the decades-old dream of “all your data at your fingertips” real.

    Nadella reflects on Bill Gates’ 1990s obsession with “information management” and structured data, noting that deep neural networks unexpectedly solved the messiness problem that rigid schemas never could.

    On bubbles: unlike the dark fiber overbuild of 2000, today Microsoft is sold out and struggling to add capacity fast enough. Demand is proven and immediate.

    On the future of work: Nadella manages by “wandering Teams channels” rather than physical halls. He stays deeply connected to startups (he visited Stripe when it was tiny) because that’s where new workloads and aesthetics are born.

    UI prediction: we’re moving toward personalized, generated IDEs for every profession – think “mission control” dashboards for orchestrating thousands of agents with micro-steering.

    Excel’s immortality: it’s Turing-complete, instantly malleable, and the most approachable programming environment ever created.

    Agentic commerce: Stripe and Microsoft are partnering to make every catalog queryable and purchasable by agents. Discovery and directed search will move almost entirely to conversational/AI interfaces.

    Company sovereignty in the AI era: the new moat is your own fine-tuned foundation model (or LoRA layer) that continually learns your tacit knowledge, combined with memory, entitlements, and tool use that stay outside the base model.

    Microsoft’s AI stack strategy: deliberately modular (infra, agent platform, horizontal & vertical Copilots) so customers can enter at any layer while still benefiting from integration when they want it.

    My Thoughts

    Two things struck me hardest:

    • Nadella is remarkably calm for someone steering a $3T+ company through the biggest platform shift in decades. There’s no triumphalism – just relentless focus on distribution inside enterprises and solving the boring data plumbing.
    • He genuinely believes the proprietary vs open debate is repeating: just as AOL/MSN lost to the open web only for Google/Facebook/App Stores to become new gatekeepers, today’s “open” foundation models will quickly sprout proprietary organizing layers (chat front-ends, agent marketplaces, vertical Copilots). The power accrues to whoever builds the best ensemble + tools + memory stack, not the raw parameter count.

    If he’s right, the winners of this cycle will be the companies that ship useful agents fastest – not necessarily the ones with the biggest training clusters. That’s excellent news for Stripe, Microsoft, and any founder-focused company that can move quickly.

  • Cloudflare Down November 18 2025: Massive Global Outage Takes X (Twitter), ChatGPT, Discord, Spotify, League of Legends & Thousands of Websites Offline

    FINAL UPDATE – Post-Mortem Released: Cloudflare has released the detailed post-mortem for the November 18 event. The outage was caused by an internal software error triggered by a database permission change, not a cyberattack[cite: 25, 26]. Below is the technical breakdown of exactly what went wrong.

    TL;DR – The Summary

    • Start Time: 11:20 UTC – Significant traffic delivery failures began immediately following a database update.
    • The Root Cause: A permission change to a ClickHouse database caused a “feature file” (used for Bot Management) to double in size due to duplicate rows[cite: 26, 27, 81].
    • The Failure: The file grew beyond a hard-coded limit (200 features) in the new “FL2” proxy engine, causing the Rust-based code to crash (panic)[cite: 190, 191, 194].
    • Resolution: 17:06 UTC – All systems fully restored (Main traffic recovered by 14:30 UTC)[cite: 32, 90].

    The Technical Details: A “Panic” in the Proxy

    The outage was a classic “cascading failure” scenario. Here is the simplified chain of events from the report:

    • The Trigger (11:05 UTC): Engineers applied a permission change to a ClickHouse database cluster to improve security. This inadvertently caused a query to return duplicate rows[cite: 160, 172].
    • The Bloat: This bad data flowed into a configuration file used by the Bot Management system, causing it to exceed its expected size[cite: 27, 125].
    • The Crash: Cloudflare’s proxy software (specifically the FL2 engine written in Rust) had a memory preallocation limit of 200 features. When the bloated file hit this limit, the code triggered a panic (specifically called Result::unwrap() on an Err value), causing the service to fail with HTTP 500 errors[cite: 190, 218, 219].
    • The Confusion: To make matters worse, Cloudflare’s external Status Page also went down (returning 504 Gateway Timeouts) due to a coincidence, leading engineers to initially suspect a massive coordinated cyberattack.

    Official Timeline (UTC)

    Time (UTC) Status Event Description
    17:06 Resolved All services resolved. Remaining long-tail services restarted and full operations restored[cite: 268].
    14:30 Remediating Main impact resolved. A known-good configuration file was manually deployed; core traffic began flowing normally [cite: 32, 268].
    13:37 Identified Engineers identified the Bot Management file as the trigger and stopped the automatic propagation of the bad file [cite: 268].
    13:05 Mitigating A bypass was implemented for Workers KV and Access to route around the failing proxy engine, reducing error rates [cite: 267].
    11:20 Outage Starts Network begins experiencing significant failures to deliver core traffic .
    11:05 Trigger Database access control change deployed[cite: 267].

    Final Thoughts

    Cloudflare’s CEO Matthew Prince was direct in the post-mortem: “We know we let you down today”[cite: 37]. The company has identified the specific code path that failed and is implementing “global kill switches” for features to prevent a single configuration file from taking down the network in the future[cite: 259].

    Read the full technical post-mortem: Cloudflare Blog: 18 November 2025 Outage