『The Insanely Great Podcast』のカバーアート

The Insanely Great Podcast

The Insanely Great Podcast

著者： Christopher John
無料で聴く

The Insanely Great Podcast is your weekly mix of tech, AI, sports, politics, and whatever other shiny object grabs our attention. One minute we’re breaking down the future of artificial intelligence, the next we’re debating game-day drama, political chaos, or the kind of random topic that somehow becomes the best conversation of the episode. Smart, funny, and just unhinged enough to keep things interesting, it’s the podcast for curious people who like their insights sharp and their conversations a little unpredictable.

2026 Christopher John 政治・政府
エピソード
  • Where did Fabel Go?

    Where did Fabel Go?
    2026/06/18

    The Insanely Great Podcast — Episode 11
    Christopher & Darrell | June 16, 2026

    The U.S. government just killed the most powerful AI model ever built — and it might be the best thing that ever happened to Anthropic. Chris and Darrell break down the sudden shutdown of Anthropic's "Fable" model, then dig into iOS 27's new AI Siri, SpaceX's trillion-dollar IPO, and a few hot takes on summer's biggest movies.

    In this episode:

    Why a government-ordered shutdown might be Anthropic's best pre-IPO marketing — and what made the model behind it uniquely dangerous

    Whether Fable users were secretly downgraded to a cheaper model the whole time

    Hands-on impressions of the iOS 27 beta and Apple's new AI Siri, including what it can actually do and where it falls flat

    Why Apple's on-device AI model is bigger than it needs to be, and what that says about the company's real priorities

    SpaceX going public, Elon Musk becoming the first trillionaire, and which parts of the business actually make money

    A review of Spielberg's "Disclosure Day," plus why Hollywood keeps losing audiences to bad sequels

    How a $200K YouTube movie made $115 million with zero CGI

    The contradiction at the heart of Apple's brand: wholesome assistant, dark prestige TV

    Whether Apple's servers can handle AI Siri once it rolls out to a billion iPhones

    The stat that 90% of OpenAI users pay nothing, and what it means for the future of AI pricing

    Plus: UFC on the White House lawn, a returning McDonald's pie, and a reflecting pool that turned green.

    New episodes every week. Find us wherever you listen to podcasts.

    ——————————————————————

    New episodes every week. Find us wherever you listen to podcasts.

    • (00:00) - Cold open & welcome
    • (00:11) - Claude's "Fable" model pulled by executive order
    • (04:03) - Was Fable quietly downgrading users to cheaper models?
    • (09:00) - Amazon, Bedrock & Mythos — AI security and zero-days
    • (11:37) - Anthropic's IPO — when a government ban becomes free marketing
    • (13:34) - iOS 27 beta first impressions & AI Siri waitlist
    • (18:32) - Apple's on-device AI — model sizes, hardware limits & privacy
    • (20:26) - AI image guardrails, jailbreaking & Chinese lab releases
    • (22:55) - Google One, YouTube Premium & subscription stacking
    • (27:02) - SpaceX IPO — Elon Musk becomes the first trillionaire
    • (29:51) - Moon bases, Mars ambitions & NASA's shrinking budget
    • (30:03) - Movie review — Spielberg's "Disclosure Day"
    • (35:07) - Movie theaters, Wonder Woman & Hollywood's sequel problem
    • (38:51) - "Obsession" — the $200K YouTuber film making $115M
    • (44:36) - Apple TV+ "Cape Fear" & Apple's brand identity
    • (46:24) - AI transcription — Grammarly, Whisper & Apple's call recording quirks
    • (48:08) - iOS 27's "Snow Leopard" moment — performance first
    • (49:05) - AI Siri deep dive — what it can actually do right now
    • (51:00) - Can Apple's servers handle Siri AI at iPhone scale?
    • (53:06) - Apple vs Google — AI model transparency & iCloud+ limits
    • (55:49) - 90% of OpenAI users don't pay — and what that means
    • (57:17) - Closing — UFC at the White House, Iran deal & a green reflecting pool
    続きを読む 一部表示
    1 時間 2 分
  • Episode 10: WWDC 2026

    Episode 10: WWDC 2026
    2026/06/09
    📝 Show NotesEpisode 10 — WWDC 2026: Faster, Smarter, and Slightly InfuriatingIt's June 8th, 2026, and Apple just wrapped WWDC — so Chris and DJ immediately got on the mic to break it all down while the keynote dust was still settling.This year's WWDC felt different from the jump. Gone was the old device-by-device structure. Instead, Apple went big on ecosystem-wide themes — and that set the tone for a keynote that was more about refinement than revelation.Here's what we got into:Liquid Glass, but make it choices. Apple walked back just enough — you can now go lighter or darker — but if you were hoping for a full "turn this off" toggle, you're still out of luck. There are accessibility workarounds floating around the internet, but Apple's not handing you the kill switch.No product announcements. Not one. Chris was genuinely surprised. WWDC used to be where MacBook Pros quietly dropped for back-to-school season. This year? Radio silence. Could a TSMC chip or memory shortage be pushing hardware reveals back to September? We think so.The Snow Leopard update is real. A lot of this keynote was Apple saying the quiet part loud: the last cycle shipped rough, so now we're fixing it. Apps launch 30% faster, AirDrop transfers are 80% quicker, there's a new CPU scheduler rolling back to iPhone 11, and the Wi-Fi-to-cellular handoff — one of iOS's longest-running frustrations — is finally getting addressed.Siri AI is here, and it's powered by Google Gemini. Worst kept secret of 2026. Craig Federighi took a very careful seat with a panel of journalists to explain that Apple isn't using Google's app, Google's infrastructure, Google's models, or Google's search. The amount of Google Assistant they use is, quote, "none." (Sure, Craig.) Chris's read: Apple took a stripped-down, whitelisted version of Gemini and parked it on their Private Cloud servers. ChatGPT isn't going anywhere, and it sounds like more AI providers are coming, which conveniently sidesteps an ongoing lawsuit about system-level AI privilege.The hardware requirements are where it gets spicy. There are two tiers. Tier 1 — standard Siri AI — runs on anything from iPhone 15 and M1 Macs up. Tier 2 — the good stuff — needs 12GB of unified memory. That means iPhone 17 Pro and up. Not the iPhone 16 Pro. Not the iPhone Air. The phone Apple literally advertised as the Apple Intelligence phone two years ago. DJ, who upgraded specifically for those AI features, had feelings about this. Chris called it what it is.The M1 Mac inconsistency. A base M1 Mac with 8GB of RAM is supported. An iPhone 16 Pro with 8GB is not. As Chris put it: that's not a technical decision, that's a marketing one. If you really want advanced AI on your iPhone and don't want to upgrade, Gemini and ChatGPT already exist as apps — and they run just fine.Photos search is finally getting smart. Ask for "Thanksgiving photos from two years ago" and the app will actually find them. Yes, Lightroom has done this for five years. But welcome to the party, Apple.macOS Golden Gate. The internal codename was Big Bear. They went with Golden Gate instead. And honestly — does anyone still care about OS names? iOS 26, macOS 26. That's it. We've evolved.Parental controls got a big segment. Chris and DJ's theory: Apple is getting ahead of legislation. With states rolling out real-ID checks for adult content and OpenAI briefly flirting with AI-based age detection, the writing is on the wall. Apple would rather build the tools than get caught flat-footed by Congress.On the beta: it's already out for developers, public beta arrives next month. The Siri AI has a waitlist because they don't want the servers crushed on day one. Chris made it to beta 8 last year before deciding to just wait the three weeks. We respect it.The Insanely Great Podcast drops new episodes whenever Apple gives us something to talk about — which, as it turns out, is pretty often. Subscribe wherever you listen.(00:00) - (00:00) - Intro (00:04) - Welcome to Episode 10 (00:40) - WWDC 2026: A Different Kind of Keynote (01:14) - Liquid Glass Gets a Refinement (Sort Of) (02:31) - Zero Product Announcements — Really? (04:57) - The Snow Leopard Update: All About Speed (05:46) - 30%, 80%, and a New CPU Scheduler (06:24) - Wi-Fi to Cellular Transitions & iMessage Fixes (07:35) - Shared Photo Galleries Go Full-Res (Finally) (08:42) - The Big News: Google Gemini Meets Siri AI (09:35) - How Apple Is Actually Running Gemini (10:19) - ChatGPT, Grok & the AI Lawsuit Drama (11:42) - On-Device AI: The Hardware Bar Is High (13:28) - Tier 1 vs. Tier 2: Who Gets What (18:39) - Photos Search Gets Smarter — For Some (20:16) - Eight Gigs on a Mac Is Fine, But Not on Your iPhone? (22:53) - macOS Golden Gate & the Death of OS Names (24:45) - Parental Controls, Age Verification & Apple Getting Ahead of Legislation (27:24) - Closing Thoughts: Excited, Betrayed, or Both? (29:46) - Should You Install the Beta? (31:22) - Goodbye & Thanks
    続きを読む 一部表示
    32 分
  • Grill, Chill & Gemini

    Grill, Chill & Gemini
    2026/05/25
    Insanely Great Podcast — Memorial Day Edition 🍔 Episode recorded: May 24, 2026 Show Notes Fire up the grill, crack open a cold one, and let Christopher and DJ keep you company while you flip burgers. This Memorial Day edition is a buffet of tech takes, hot rants, and pop culture detours — perfect for that backyard hang. On the menu this week: 🤖 Google I/O lays the smackdown. Gemini 3.5 Flash now allegedly outguns 3.1 Pro at half the power, Gemini Spark is the big consumer cloud play, and we got a sneak peek at what Siri could be if Apple ever caught up. Spoiler: Apple is paying billions for a whitelisted, watered-down Gemini and the guys aren't sure why. 👓 The smart glasses arms race. Meta's Ray-Bans, Oakleys, Apple's incoming pair, and now Google's at the party too. DJ returned his Oakleys (vertical video — really, Meta?) and Christopher's Ray-Bans live on the shelf 90% of the time. Four-hour battery life is, in technical terms, "abysmal." 🏗️ Android 17 "Cinnamon Bun" gets deeper Gemini integration. In other shocking news, water is wet. 🛠️ Google Antigravity and the slow death of the IDE — just tell the AI what to code and walk away. 🍌 Nano Banana 2 keeps being magical, and Christopher demands to know how these models change one thing without nuking the whole image. 🚀 SpaceX files for what could be the biggest IPO in history (ticker: SPCX). Also, hey, at least the toilet didn't explode this time. 💸 NVIDIA prints money ($91B quarter, $80B buyback) while we debate whether Anthropic could ever catch the shovel-seller. 🏢 Meta's layoffs hit 8,000 while employees literally train their AI replacements via keystroke surveillance. Salesforce pulled the same gaslight. Yikes. 🧠 The AGI debate gets philosophical. Are LLMs the steam engine of intelligence? Plus a fascinating tangent on world models, three-system AI learning, and why power-hungry "evolutionary" AI might actually get us there. 🎬 Mandalorian & Grogu is a two-and-a-half-hour nothing burger. Sigourney Weaver: why are you here? Where are Luke, Ahsoka, Han, and Leia? A huge missed opportunity gets thoroughly dragged. 🪄 Harry Potter marathon! DJ is initiating his 13-year-old into Hogwarts (he's never seen them!). Christopher swears by the Chris Columbus extended editions, dishes on the upcoming HBO Max reboot, and yes — they get into the Hagrid actor situation. 💍 Bonus rant: Peter Jackson recording 30-minute intros to already-four-hour extended LOTR movies. The nerds applauded. Of course they did. 🎮 Gaming corner: Diablo 4's Mephisto expansion, the new James Bond game (looks like 2005?), Hogwarts Legacy love, Crimson Desert gunning for Game of the Year, the underwhelming Switch 2 launch lineup, and rumors of a $1,500 Xbox PC-console hybrid. 📺 YouTube Premium price hike gets roasted. $26/month for a family plan? For… ad-free and downloads? Bundle it with YouTube TV or stop, Google. 📱 iPhone Ultra fold rant preview (saved for next episode): how do you call it "Ultra" when the camera lenses are worse than the Pro? 🚫 Apple Shortcuts as moral police. Try to get it to proofread anything with the word "ass" and watch it silently refuse to do its job. Cool, cool, cool. 🔥 DJ's pro tips: DeepSeek v4 Flash in Hermes agent has been ripping, and check out Ion Router — they're running the OSS 120B model at 5¢ in / 10¢ out by squeezing NVIDIA GH200s. Insanely cheap. Let us know if you try it. 🎓 And a sentimental dad moment — congrats to Christopher's daughter on graduating high school yesterday. Sniff. Happy Memorial Day, everyone. Hug your veterans, hug your data centers, and we'll see you next week. 📧 Got thoughts on Ion Router or anything else? Hit us up — we read everything. Chapter Markers (00:00) - Welcome & Memorial Day Burger Talk (00:39) - Google I/O: Smaller Agents Are the Future (01:22) - Gemini 3.5 Flash Beats 3.1 Pro (Allegedly) (02:37) - The Siri + Gemini WWDC Mystery (04:31) - Why Apple Gets a Watered-Down Gemini (06:18) - Smart Glasses Are Coming for All of Us (07:06) - Christopher's Ray-Bans Live in a Drawer (07:59) - DJ's Oakley Return Saga (Vertical Video?!) (09:14) - The End of Traditional Search (10:18) - Android 17 "Cinnamon Bun" — Surprise, More Gemini (10:50) - Google Antigravity & Vibe-Coding IDEs (12:09) - Nano Banana 2 Is Witchcraft (13:02) - SpaceX Files for IPO (Ticker: SPCX) (14:25) - NVIDIA Earnings Blowout & the China Chip Problem (17:27) - Meta Layoffs: 8,000 Cut, Employees Training Their Replacements (20:47) - Trump's AI Executive Order Delayed (22:09) - The AGI Debate: Are LLMs the Steam Engine? (23:25) - World Models & Three-System AI Learning (25:22) - Mandalorian & Grogu Movie Review (Oof) (29:37) - Harry Potter Marathon Time (30:35) - The Case for the Chris Columbus Extended Editions (32:48) - Peter Jackson's 30-Minute LOTR Intros (Bless Him) (34:47) - Star Wars vs. Star Trek vs. Lord of the Rings Geek Cultures (35:17) - The Janky-Looking ...
    続きを読む 一部表示
    55 分
adbl_web_anon_alc_button_suppression_t1
まだレビューはありません