Leo Laporte opens TWiT episode 1086 with an energetic teaser running through the week's most compelling stories: a spectacular rocket explosion, the Pope's AI encyclical, Ferrari's first EV, and a fake legal agreement buried forever in the Bitcoin blockchain. He introduces his three guests with characteristic warmth — Gary Rivlin, a Pulitzer Prize-winning journalist whose book AI Valley covers the trillion-dollar AI race; Molly White, creator of Web3 Is Going Just Great and the Citation Needed newsletter; and Sam Abuelsamid, VP of Research at Telemetry Agency and co-host of Wheel Bearings. The rapid-fire intro sets the tone for a wide-ranging, intellectually serious panel.

With SpaceX's IPO S-1 still fresh off the press, Gary Rivlin leads a sharp dissection of its most audacious provisions. The standout: Elon Musk retains voting rights on performance shares that haven't been granted yet, contingent on benchmarks like putting a million people on Mars — meaning he can vote equity he doesn't own. ^{[1] — Gary Rivlin "SpaceX's S-1 IPO filing contains a provision allowing Elon Musk to vote performance shares that haven't been granted to him yet — shares co…" 03:30 Play} Rivlin also highlights a $26.5 trillion actionable total addressable market for AI that he flatly calls fabricated. Sam Abuelsamid adds a crucial economic reality check: strip out Starlink and SpaceX's actual rocket-launch business loses money even after receiving roughly $38 billion in federal government revenue. The group notes that SpaceX's corporate governance would be scandalous at any conventional public company, but investors will likely pile in anyway, betting on the Musk halo effect rather than the underlying financials — treating it as a meme stock before it's even listed.

The most spectacular rocket explosion since the Soviet N1 in 1969 has just occurred in Florida, and Sam Abuelsamid walks the panel through exactly what happened: Blue Origin's New Glenn was undergoing a hot fire test — engines ignited while the vehicle remains clamped to the pad — when it exploded, destroying the facility that is the company's only New Glenn launch site. ^{[1] — Sam Abuelsamid "Blue Origin's New Glenn rocket exploded during a hot fire test, destroying the company's only launchpad and setting back NASA's lunar progr…" 07:33 Play} That means at least a year of rebuild time, directly threatening NASA's timeline for launching lunar rovers. Abuelsamid widens the lens to critique the environmental cost of the methane rocket era: liquid methane boils off continuously from cryogenic tanks even sitting on the pad, venting a greenhouse gas far more potent than CO2. Molly White sharpens the cultural critique, noting that Silicon Valley's 'move fast and break things' philosophy is now being applied to industries — space, government, critical infrastructure — where that ethos has genuinely catastrophic consequences.

Leo Laporte transitions to a Mill food recycler sponsorship segment that feels more like a product review than an ad: he pulls up the Mill app live on air and shows the panel that his household has kept 413 pounds of food waste out of landfills since purchasing the device in October. The conversation quickly becomes a genuine panel discussion as Sam Abuelsamid reveals he has been composting for 30 years using a 55-gallon tumbler composter, and Molly White discusses the challenges of home composting for city dwellers. Leo explains that Mill quietly dehydrates and stirs food scraps overnight, reducing them to coffee-ground-like material that can be used in gardens or returned to Mill for small farm use. The segment includes practical tips — tea bag strings jam the stirrer, coffee grounds are fine, and even chicken bones can go in — before Leo wraps with the offer: $75 off at mill.com/twit with code TWIT.

Leo Laporte pivots to what he considers one of the week's most intellectually interesting stories: Pope Leo XIV — who named himself after the Pope who wrote a landmark encyclical about the Industrial Revolution — has released Magnifica Humanitas, a 42,000-word document on AI. Gary Rivlin is enthusiastic: the Vatican has been sending delegations to Silicon Valley since the mid-2010s, meeting with Reid Hoffman, Google's James Manyika, and Microsoft's Kevin Scott, and the resulting document reflects genuine depth. ^{[1] — Gary Rivlin "Pope Leo XIV's first encyclical is a surprisingly sophisticated AI policy document, built on years of Vatican engagement with Silicon Valle…" 41:30 Play} Rivlin's summary of its core argument: AI is neither inherently good nor bad — the problem is who controls it, and a handful of tech elites in Silicon Valley and Beijing should not be making those decisions for all of humanity. Molly White adds what she finds most valuable: the Pope explicitly challenges the tech industry's 'AI is inevitable' framing, arguing that AI is a human technology that humans can choose how — or whether — to develop. The panel notes that the encyclical is pro-AI in the sense that it acknowledges AI's genuine benefits, while calling for guardrails, human oversight, and democratic accountability for its development.

Gary Rivlin opens with a lament: in 2023, the United States was having a genuine national conversation about AI regulation, with Sam Altman publicly asking Congress to regulate his industry. Then it went 'poof.' The trillion-dollar race took over, China became the all-purpose excuse for inaction, and safety concerns were discarded. Molly White adds an important nuance: Trump's promise to the AI and crypto industries wasn't 'no regulation' but rather 'you write the regulation,' ^{[1] — Molly White "Hundreds of Wikipedia volunteer editors are threatening to strike after the Wikimedia Foundation disbanded its dedicated community wishlist…" 2:22:27 Play} enabling incumbents to use compliance costs as a competitive moat against new entrants — exactly the pattern she has tracked in the crypto industry. The panel discusses Trump's aborted executive order that would have required government approval of AI model releases before Anthropic's controversial Mythos model spooked the White House, noting that David Sachs reportedly led an 11th-hour call that killed the order. The fundamental tension: approving models is too much government overreach, but having no guardrails enables harm. Nobody in government seems to know what good AI regulation even looks like.

Leo Laporte delivers a Box sponsorship segment centered on a problem he finds genuinely compelling: AI models are powerful on public knowledge but blind to the specific institutional content — product roadmaps, HR policies, financial models — that actually runs a business. Box, he explains, solves this by serving as a secure context layer that gives AI agents access to enterprise content without exposing it publicly. He highlights Box Agent, Box Extract, and Box Hubs as tools that let organizations turn unstructured content into actionable intelligence, and emphasizes Box's security-first approach: employees and agents only access information they're authorized to see. The URL box.com/ai is the call to action.

Molly White drops a significant announcement: she is expanding her Follow the Crypto election spending tracker — which has documented crypto industry's massive political influence — to cover AI lobbying as well, renaming it Tech Influence Watch, available via citationneeded.news. ^{[1] — Molly White "Hundreds of Wikipedia volunteer editors are threatening to strike after the Wikimedia Foundation disbanded its dedicated community wishlist…" 2:22:27 Play} Gary Rivlin makes the stunning observation that Andreessen Horowitz has become the single largest political donor in the current election cycle, outspending even Elon Musk's $250 million and eclipsing traditional power industries like oil and banking. The panel details specific examples: AI and crypto PACs spending tens of millions against California state legislators who proposed mild AI regulations, Palantir funding attack mailers against a data scientist state legislator, and crypto groups running attack ads against politicians while the super PAC's own donors have ICE contracts. Molly White draws the connection to Citizens United: when the payoff of spending $100 million on lobbying is billions in favorable regulation, the math always works out in favor of spending. She notes a rare bipartisan note of optimism — Maine's 2024 referendum to limit corporate political spending passed with 75% of the vote.

Sam Abuelsamid offers what may be the episode's most practically grounded insight: the auto industry lobbied vigorously against seatbelts, emissions standards, and every major safety regulation since the 1960s — and lost. In the 60 years since, the industry has produced more innovation than in its entire prior 60-year history. Cars are safer, more efficient, and more capable than ever. ^{[1] — Sam Abuelsamid "The auto industry lobbied furiously against seatbelts and emissions standards. They lost. And in the 60 years since, the industry innovated…" 50:00 Play} The constraints forced creativity. He uses this to directly challenge the tech industry's claim that AI regulation would kill innovation. Gary Rivlin extends the historical analogy: railroads became trustworthy — and thus enormously profitable — only after government mandated standardized track gauges and block signaling systems. Molly White adds that American financial markets are trusted globally precisely because of regulation, and that the erosion of that regulation under the current administration is visibly damaging that trust. Leo Laporte acknowledges the hard problem: it's easy to say a car needs a seatbelt, but it's not clear what 'make an AI safer' even means technically.

Leo Laporte opens the Zscaler segment by cataloguing the security threats that AI has amplified: AI-generated phishing lures indistinguishable from real communications, AI-assisted malicious code writing, AI-accelerated data exfiltration, and employees unwittingly uploading sensitive company information to public AI models — including 1.3 million Social Security numbers leaked to AI apps last year. Zscaler's Zero Trust architecture removes attack surfaces by default while its AI layer monitors how employees use generative AI tools, preventing sensitive data from reaching public LLMs. A testimonial from Siva, Director of Security and Infrastructure at Zwoora, explains how Zscaler ZIA gives visibility into employee AI tool usage and enforces data classification policies. The call to action is zscaler.com/security.

Google has gone all-in on AI Overviews, placing AI-generated answers at the top of every search result and giving users no easy way to opt back to blue links. The market response is immediate: DuckDuckGo installs jumped 30% in a single week, ^{[1] — Leo Laporte "Google has bet its search product on AI Overviews, putting AI-generated answers front and center. The backlash is immediate: DuckDuckGo ins…" 1:13:55 Play} with Molly White reporting her own mother discovered DuckDuckGo independently. Gary Rivlin frames the deeper problem: with approximately 58% lower click-through rates to source websites from AI search, the financial model sustaining the web's content creators is collapsing. Leo Laporte confesses he has already largely stopped using search engines — his AI agents query X, Reddit, Hacker News, and his RSS feeds directly, synthesizing the top stories. The panel identifies the ouroboros problem Molly White has tracked in research: AI scrapes the web, generates AI content, the web becomes AI slop, future AI trains on that slop, and model quality degrades through 'model collapse.' Gary Rivlin quotes Mustafa Suleiman acknowledging that Microsoft is trying to avoid training on AI-generated data — but admits the distinction is increasingly hard to enforce.

The philosophical heart of the episode's AI discussion: can AI ever truly replace human creative content? Gary Rivlin draws on his experience finding AI-generated podcasts analyzing his own book — impressive in their way but unmistakably hollow. ^{[1] — Gary Rivlin "AI is obviously artificial intelligence, but I think it's alien intelligence. It knows everything and understands nothing. It has no common…" 1:24:25 Play} He argues that what journalists, analysts, and podcasters actually sell is not information but curation: taste, judgment, ethics, and instinct that no model trained on averages can replicate. Molly White notes that AI companies underestimate how much people value human-made material — music, newsletters, journalism — and that a significant consumer backlash is already underway, with teenagers carrying CD players as a statement against digital everything. Gary Rivlin invokes the photography analogy: when cameras arrived, people said painting was dead; instead, painting became a more valued artform. The youngest generation, he observes, is actually the most skeptical of AI — they're the ones booing tech executives at commencement ceremonies because AI is coming for their entire working lives.

Leo Laporte delivers ZipRecruiter's sponsorship segment with a focus on the platform's ability to surface not just qualified candidates but specifically those who are most enthusiastic about the role — candidates who explain in their own words why they want the job. He notes ZipRecruiter is rated the number one hiring site on G2, with four out of five employers receiving a quality candidate within the first day of posting. Listeners can try it free at ziprecruiter.com/twit.

The panel digs into the question of whether AI will ever match human literary creativity. Gary Rivlin makes his most memorable argument of the episode: AI is not artificial intelligence, it's alien intelligence — it has ingested the entirety of human creative output but possesses no common sense, no lived experience, and no understanding. ^{[1] — Gary Rivlin "AI is obviously artificial intelligence, but I think it's alien intelligence. It knows everything and understands nothing. It has no common…" 1:24:25 Play} He uses the New York Times's Curtis Sittenfeld experiment — same prompt given to a celebrated novelist and to an AI — to argue that great writing is immediately recognizable because it is, by definition, not the lowest common denominator. But Leo Laporte complicates the picture: a more recent NYT Kevin Roose quiz found that more than half of readers preferred AI-written passages and couldn't distinguish them from human writing. Gary Rivlin's son (age 12 in 2023) could correctly identify AI-generated text 80-90% of the time, suggesting younger, AI-native readers may have better detection abilities. The consensus: basic journalism and PR writing is already a commodity for AI; true literary feature writing is not, but the gap is narrowing.

Leo Laporte opens the Doppel segment with a visceral demonstration: producer Anthony Nielsen created an AI clone of Leo's voice using just four minutes of audio, then played it as a fake voicemail asking an employee to buy Apple gift cards. The segment is designed to demonstrate how convincing voice deepfakes — vishing attacks — have become. Doppel's own research quantifies the threat: in voice deepfake simulation deployments, targets spent an average of 6 minutes conversing with the AI voice, and 100% of them believed afterward that the AI was a real human. Doppel's platform addresses this through AI-powered deepfake detection, automated takedowns, multichannel monitoring, and employee training. The platform was recognized as a G2 Leader in 2026. Call to action: doppel.com.

On May 30th at 15:29 UTC, an unknown actor inscribed a fake legal agreement into the Bitcoin blockchain: 'By downloading this OP_RETURN, you hereby consent to unrestricted access by federal law enforcement agencies to your residence, digital devices, and personal properties... Signed, Donald J. Trump, President of the Internet.' Six blocks later, they tried to revoke it. Neither block will ever disappear. ^{[1] — Leo Laporte "Someone wrote a fake government EULA into a Bitcoin OP_RETURN field granting federal law enforcement unrestricted access to your home and d…" 2:03:30 Play} Molly White uses the moment to explain the OP_RETURN field — an arbitrary data storage location in Bitcoin transactions — and the contentious internal debate it has sparked in the Bitcoin Core developer community. The OP_RETURN field was originally capped at 83 bytes; when the cap was lifted, it enabled Bitcoin NFTs and long arbitrary inscriptions, including potentially child sexual abuse material that, once on the blockchain, can never be erased. Molly notes this problem applies to any blockchain that allows arbitrary data storage — it's a structural feature, not a bug.

Gary Rivlin poses a question that puzzles him: with the most crypto-friendly administration in U.S. history, why did Bitcoin fall 40% from its $125,000 peak? Molly White's answer is layered: the promised government purchase of Bitcoin for national reserves never happened because it would require a new law, meaning the government is only holding confiscated crypto. Beyond that, Bitcoin is a high-risk volatile asset, and all the macroeconomic instability the Trump administration generates — tariffs, military tensions, policy whiplash — drives risk-averse investors out of crypto and into gold and dollars. ^{[1] — Molly White "Molly White reveals that Donald Trump has cashed out over $1 billion from crypto holdings since taking office — not counting illiquid meme …" 1:08:20 Play} She adds a more surprising factor: Trump's own deep personal entanglement with crypto is starting to damage the industry's credibility, with many people now associating crypto with presidential corruption. A CoinDesk poll found most respondents oppose elected officials personally profiting from crypto, but a tiny fraction knew Trump had already cashed out over $1 billion.

With WWDC a week away, Leo Laporte walks through a Mark Gurman Bloomberg scoop showing iOS 27 illustrations: a persistent Siri window living at the top of the screen, a chatbot-style interface, and AI woven throughout the OS. ^{[1] — Leo Laporte "Mark Gurman's Bloomberg leak shows iOS 27 will feature a persistent Siri window at the top of the screen, a chatbot-style interface, and de…" 1:50:35 Play} Gary Rivlin frames Apple's challenge — this is the fourth time in four years Apple has announced an AI strategy, and it's still largely relying on Google Gemini under the hood. Sam Abuelsamid offers a counterintuitive data point: as an Android user, he has found Gemini significantly less functional than the Google Assistant it replaced, citing broken integrations with smart home devices like garage door openers and LED lights. Apple's WWDC moment will matter, the panel agrees, not because of the technology itself but because putting a polished AI interface in front of a billion iPhone users is a different kind of milestone than anything Google or Microsoft has done. Molly White notes that AI is already ubiquitous in Gmail, Google Search, and every major app — one more insertion may not be the sea change Leo expects, though Apple's historically gentler touch could make a difference.

Leo Laporte offers a peek behind the curtain of his personal AI agent setup: a tool from Nuance Research called Hermes, which he has named Quicksilver. Unlike Siri or Alexa, Quicksilver maintains persistent memory of prior conversations, knows Leo's dietary habits, tracks his meals, and queries X, Reddit, Hacker News, and RSS feeds on command to surface the week's top AI stories. Gary Rivlin agrees the personal assistant is the holy grail — and that it's still 2-3 years away from mainstream deployment, placing true 'year of the AI agent' somewhere in 2027 or 2028. ^{[1] — Leo Laporte "Mark Gurman's Bloomberg leak shows iOS 27 will feature a persistent Siri window at the top of the screen, a chatbot-style interface, and de…" 1:50:35 Play} Sam Abuelsamid pushes back on the sassy voice assistant trend: he doesn't need Alexa Plus to make jokes about his asparagus timer; he needs it to open his garage door, which Gemini still can't do. Molly White adds that she fundamentally dislikes talking to computers and doubts she'll get over it. The panel briefly explores subvocalization interfaces and EEG headsets as potential alternatives for office environments where speaking to a computer is socially awkward.

Molly White walks through the Wikipedia crisis with the authority of a longtime insider. The community wishlist is the mechanism by which Wikipedia's volunteer editors request software features from the Wikimedia Foundation's paid engineering staff. The Foundation recently announced it was eliminating the dedicated team that managed this work, claiming the change would actually improve responsiveness by distributing the work across all engineers. ^{[1] — Molly White "Hundreds of Wikipedia volunteer editors are threatening to strike after the Wikimedia Foundation disbanded its dedicated community wishlist…" 2:22:27 Play} The volunteer community rejected that framing entirely, seeing it as deprioritization — not improvement. The situation was further inflamed by allegations that some of the laid-off engineers had been active in an ongoing effort to unionize Wikimedia Foundation employees, suggesting the layoffs may have been retaliatory union-busting. Molly notes she is personally a member of a committee specifically designed to improve technical communication between the Foundation and the community, and that even this committee was barely consulted. The Foundation's public response was deemed insufficient. The episode sits within a decades-long pattern of the Foundation stepping on rakes in its relationship with the volunteer community, despite — or perhaps because of — those volunteers being the irreplaceable foundation of everything Wikipedia is.

Leo Laporte presses on a question with major implications for the AI industry: do AI companies support Wikimedia financially? Molly White explains the Wikimedia Enterprise model: AI companies pay not for the data itself but for a premium API — a fast, well-formatted access channel that makes ingestion far easier than scraping. She considers this a reasonable arrangement. The controversy, she recalls, arose partly from confusingly written headlines that made it sound like Wikipedia was incorporating AI-generated content into the encyclopedia — which it is not. Leo notes the irony that if Wikipedia's volunteer editor community follows through on a strike, it would not just damage Wikipedia but potentially every major AI model trained on its data. Gary Rivlin closes with an observation that captures Wikipedia's remarkable arc: he worked at the New York Times in the mid-2000s when using Wikipedia as a source was a fireable offense, and now it is arguably the single most important training resource for the AI systems that will shape civilization.

Sam Abuelsamid holds court on the Ferrari Luce, the Italian automaker's first fully electric vehicle, co-designed by Jony Ive's studio LoveFrom. His verdict is nuanced: the interior has genuine merit — an analog clock, a pivoting touchscreen, and manual airflow vents that don't require you to dig through software menus — but the exterior simply doesn't read as a Ferrari. It could pass for a Honda Zero Series concept or a Jaguar I-Pace. ^{[1] — Sam Abuelsamid "Ferrari's first EV, the Luce, costs $640,000 and carries design input from Jony Ive's firm LoveFrom. Sam Abuelsamid's verdict: the interior…" 2:41:20 Play} Sam contextualizes this within Ferrari's broader strategic logic: like Porsche with the Cayenne SUV (now its best-selling model), Ferrari needs vehicles beyond sports cars to survive. Without Porsche's Cayenne and Panamera, the 911 would not be economically viable today. The $640,000 price, the design pedigree, and Jony Ive's involvement make this a cultural moment, but not necessarily a successful one. The panel lands on what everyone has been thinking: this looks like what Apple's Project Titan car would have looked like, designed for a self-driving interior that was never going to exist. Leo adds that Ive is now designing an OpenAI device — rumored to be a pendant — due sometime next year.

The New York Times reports Peter Thiel — PayPal co-founder, Palantir investor, and one of the architects of Silicon Valley's political influence — has relocated himself and his family to Argentina, enrolled his children in local schools, and is being offered permanent residency or citizenship by the Argentine government. He has also purchased properties in the 'Beverly Hills of Buenos Aires' and neighboring Uruguay. ^{[1] — Leo Laporte "Peter Thiel, one of the architects of Silicon Valley's influence over American politics through companies like Palantir, has relocated to A…" 2:27:00 Play} Gary Rivlin, who profiled Thiel in the New York Times Magazine in 2005, recalls Thiel in the 2010s talking about leaving the U.S., about 'twisting the kaleidoscope' to break what he saw as a broken system. The panel is darkly amused: Thiel helped create the conditions he's now fleeing. Molly White notes that tech billionaires accumulating passports is a pattern she watches — citing Douglas Rushkoff's book Survival of the Richest about billionaires building bunkers because they fear what they've created. Sam Abuelsamid offers a practical suggestion: load them all on a Starship and send them to Mars, which is where they seem to want to go anyway.

As the Wall Street Journal reports that robotaxis are spreading across the U.S. — along with the backlash — Sam Abuelsamid walks through the specific incident log: multiple Waymos drove into post-storm floodwaters in Atlanta with passengers aboard. A string of Waymos became stuck circling a neighborhood roundabout endlessly, like army ants caught in a spiral. There are ongoing failures in recognizing school buses and responding correctly. Waymo has paused all highway operations and reduced service in multiple cities while troubleshooting. ^{[1] — Sam Abuelsamid "Waymo robotaxis drove into floodwaters with passengers, got stuck circling roundabouts endlessly, and failed to recognize school buses. The…" 2:31:05 Play} A core technical problem: Waymo's transition over the past 8-9 years from a rules-based system to a largely AI-based monolithic software stack makes it fundamentally harder to diagnose what's going wrong. Morgan Stanley projects 30% of the rideshare industry will be autonomous by 2032, but Sam is skeptical — the economics still don't work, which is why Ford and GM have both exited the robotaxi business. The near-term model, Sam predicts, is robotaxis providing baseline service during low-demand periods while human drivers supplement during peaks.

Leo Laporte highlights Shift, a startup offering free home cleaning in exchange for first-person video training data for household robots. The panel is skeptical. Sam Abuelsamid delivers the most deflating piece of information in the robotics segment: Joanna Stern's recent Wall Street Journal video exposé found that most of the 'autonomous' household bipedal robots being demonstrated are actually being teleoperated — a human sitting in Silicon Valley watches a camera feed through the robot's eyes and controls its movements remotely. The only robots that are genuinely impressive without teleoperation, Sam says, are Boston Dynamics' Atlas and a handful of Chinese robots. Molly White admits she has converted to robot vacuums, particularly praising a Dreame model for its mopping capability, while noting her dog is merely puzzled rather than threatened by it.

Sam Abuelsamid uses the final panel segment to announce Operation Frodo's latest drive, named after the Lord of the Rings hero. Starting June 7th, a fleet of automotive media journalists will transport 12-13 rescue beagles from Omaha, Nebraska to Portland, Oregon — a 2,700-mile journey covering Cheyenne, Salt Lake City, and Ontario, Oregon — in an entirely electric vehicle fleet for the first time. ^{[1] — Sam Abuelsamid "Sam Abuelsamid's Operation Frodo charity drive transports rescue beagles from Nebraska to the Pacific Northwest, where demand for the breed…" 2:49:20 Play} The vehicles: a Cadillac Escalade IQ, Lucid Gravity, Hyundai Ioniq 9, and Kia EV9, all loaned by the manufacturers. EVgo is sponsoring free charging at Pilot Flying J stations along I-80. Sam explains the heartbreaking reason there is a surplus of beagles in the Midwest each fall: hunters abandon or shoot dogs that don't perform at the end of hunting season, and puppy mills dump unsold pups at around 4 months old. Meanwhile, demand for beagles in the Pacific Northwest far exceeds supply. Operation Frodo has now transported 88 dogs over four annual drives. The long-term goal: purchase and equip dedicated rescue transport vehicles for shelters and rescue organizations nationwide. Donations accepted at Animal Rescue Rigs. Leo immediately wants one dropped off in Petaluma.

Leo Laporte wraps the show with individual plugs for each guest: garyrivlin.com for Gary Rivlin's AI Valley and future projects; citationneeded.news for Molly White's newsletter, podcast, and upcoming Citizens United episode; wheelbearings.media for Sam Abuelsamid's car podcast. Sam gets one final plug for Operation Frodo on Animal Rescue Rigs. Leo gives the full TWiT listener guide — the show streams 2-5 PM Pacific every Sunday on YouTube, Twitch, X, Facebook, LinkedIn, and Kick; audio and video downloads at twit.tv; Club TWiT at twit.tv/clubtwit for ad-free feeds and member Discord. He closes with the show's 21-year-old sign-off: 'Another TWiT is in the can.' A post-roll Mint Mobile sponsorship ad featuring Ryan Reynolds' voice follows.