エピソード

  • Episode 279: Live from Room 101
    2025/12/01
    This week Dave (https://dgshow.org/hosts/dave) and Gunnar (https://dgshow.org/hosts/gunnar) go on a re-education off-site, and O’Brien and Winston Smith debate whether we’re in an AI bubble! Key Takeaways (The Correct Perspective): * The concept of an 'AI bubble' is a persistent, troubling rumor. It is heresy. * Nvidia's valuation is not speculation; it is the correct price for the tools that build order. * The true return on investment for General AI is not measured in commerce or email-writing, but in compliance. * The massive power consumption and centralization of the technology are not a bug, but a feature—it ensures only the Party can wield true AI. * The true 'killer app' of AI is the ability to rewrite the past perfectly and instantly, changing not just the record, but the memory. * A bubble is fragile. This new architecture of control is not a bubble, but the AI reality. * The numbers are meaningless. The only thing that matters is the fullness and the order. Further Re-education: * All corrected notes and the transcript that has now always been true can be found at dgshow.org (http://dgshow.org). We Give Thanks * Big Brother * O’Brien and Winston Smith for filling in for us!
    続きを読む 一部表示
    10 分
  • Episode 278: I Know How the Butlerian Jihad Works
    2025/11/24
    This week Dave (https://dgshow.org/hosts/dave) and Gunnar (https://dgshow.org/hosts/gunnar) talk about weaponized AI and seemingly conscious AI! Virtual Customer Panels with Gemini Gems (https://gemini.google/overview/gems/) Dave vs. Ticketmaster (https://www.ticketmaster.com/) Totally unrelated: Cool Dad Raising Daughter On Media That Will Put Her Entirely Out Of Touch With Her Generation (https://theonion.com/cool-dad-raising-daughter-on-media-that-will-put-her-en-1819572981/) Weaponizing image scaling against production AI systems (https://blog.trailofbits.com/2025/08/21/weaponizing-image-scaling-against-production-ai-systems/) Anthropic's Claude stocked a fridge with metal cubes when it was put in charge of a snacks business (https://www.engadget.com/ai/anthropics-claude-stocked-a-fridge-with-metal-cubes-when-it-was-put-in-charge-of-a-snacks-business-162750304.html) Seemingly Conscious AI (SCAI) We must build AI for people, not to be a person (https://mustafa-suleyman.ai/seemingly-conscious-ai-is-coming) Microsoft’s AI Chief Says Machine Consciousness Is an ‘Illusion’ (https://www.wired.com/story/microsofts-ai-chief-says-machine-consciousness-is-an-illusion/) Ohio lawmaker proposes comprehensive ban on marrying AI systems and granting legal personhood (https://www.msn.com/en-us/news/other/ohio-lawmaker-proposes-comprehensive-ban-on-marrying-ai-systems-and-granting-legal-personhood/ar-AA1OZhaG) Cutting Room Floor Hosting a WebSite on a Disposable Vape (https://bogdanthegeek.github.io/blog/projects/vapeserver/) How Many Phones Sport A 5 And 1/4 Diskette Drive? This One. (https://hackaday.com/2025/09/27/how-many-phones-sport-a-5-and-1-4-diskette-drive-this-one/) 1990 Airstream NASA 025 Command Vehicle. A Once-in-a-Lifetime Investment Opportunity (https://www.hemmings.com/listing/1990-airstream-all-models-478166) Steamed Hams but it's a Critically Acclaimed Feature Film (https://www.youtube.com/watch?v=pk-Oq8iYtVA) See also: Steamed Hams but it was banned in the USSR (https://www.youtube.com/watch?v=yVmw3ZhdzEs) Steamed Hams but it's a German Expressionist Film (https://www.youtube.com/watch?v=Jsof0cwFPY0) Steamed Hams but it's in Nazi Germany (https://www.youtube.com/watch?v=zY17_nraW68) The Simpsons - Worker and Parasite (https://www.youtube.com/watch?v=z2_dhUv_CrI) Possible inspiration for Worker and Parasite? (https://www.reddit.com/r/TheSimpsons/comments/x9ze4c/possible_inspiration_for_worker_and_parasite/)
    続きを読む 一部表示
    37 分
  • Episode 277: Mission Accomplished
    2025/11/17
    This week Dave (https://dgshow.org/hosts/dave) and Gunnar (https://dgshow.org/hosts/gunnar) are joined by Dan Walsh (https://dgshow.org/guests/dan) on the eve of his retirement! What is new in SUSE Linux Enterprise Server 16.0 (https://www.suse.com/c/what-is-new-in-suse-linux-enterprise-server-16-0/): “The most significant security update is the transition from AppArmor to SELinux as the default Mandatory Access Control (MAC) framework.” Bootc (https://developers.redhat.com/articles/2024/09/24/bootc-getting-started-bootable-containers) Dan Walsh Red Hat Obituary (https://www.youtube.com/watch?v=W8AATlOq3SU) SELinux (https://en.wikipedia.org/wiki/Security-Enhanced_Linux) Dan’s Russian Blog (http://danwalsh.livejournal.com) Cursor (https://cursor.com/) Podman In Action (https://www.amazon.com/Podman-Action-generation-container-engines/dp/1633439682) RamaLama (https://ramalama.ai/) vLLM (https://github.com/vllm-project/vllm) Llama.cpp (https://github.com/ggml-org/llama.cpp) Red Hat Lightspeed (https://docs.redhat.com/en/documentation/red_hat_lightspeed/) Multimodal learning (https://en.wikipedia.org/wiki/Multimodal_learning) SELinux Coloring Book (https://people.redhat.com/duffy/selinux/selinux-coloring-book_A4-Stapled.pdf) “Not for Sharing, Kindergarten fan club” (https://photos.app.goo.gl/cY22o2P3JHLZJBuY8) We Give Thanks Dan Walsh (https://dgshow.org/guests/dan) for being our special guest star! Special Guest: Dan Walsh.
    続きを読む 一部表示
    57 分
  • Episode 276: I Think I Had That Shirt
    2025/11/01
    For our Halloween finale, Carolyn Ford (https://dgshow.org/guests/cford) invites two favorite “AI conjurers”, Dave Egts (https://dgshow.org/hosts/dave) “Mr. X,” public-sector field CTO at Mulesoft, and Laura Klebanow, founder & chief storyteller at Show & Tell, into the Tech Transforms cauldron. Each brings a “secret AI ingredient”: Carolyn composes whimsical Suno-generated theme music (after a prompt-crafting detour through ChatGPT), Laura pushes image creation and Gemini’s career-planner for playful, practical use, and Dave reimagines Gemini’s children’s-storybook gem to craft customer-hero narratives, plus a Scooby-Doo-style Tech Transforms tale. Then we stir in the shadows: misinformation at scale, AI “rights,” bioethics (organoids/bodyoids), job disruption vs. reskilling, and how creativity can demystify AI without sugarcoating risks. The trio closes with rapid-fire “treats & tricks,” classroom-to-boardroom starter tips (talk with AI, don’t just query it; red-team your ideas), and a reminder to “find your beautiful” use the tools to clear the junk so humans can do what only humans do. Brewed takeaways: * Prompt craft matters; collaborate across tools (Suno, Gemini, ChatGPT). * Creative play ≠ fluff: it’s a safe on-ramp to literacy and adoption. * AI will change jobs; map “jobs to be done,” automate the rote, upskill the rest. * Ethics aren’t optional: teach harm awareness early (mis/dis/malinformation, bullying). * Start small today: keep a running dialogue with AI and build a personal “advisory board” of model personas. Show links: * Laura Klebanow: Email (mailto:laura@showtell.io) and LinkedIn (https://www.linkedin.com/in/lauraklebanow/) * Suno (https://suno.com/) * Carolyn's storybook (https://g.co/gemini/share/56bbf0498a84) * Dave's storybook (https://gemini.google.com/share/7eb18718530b) * Tech Transforms Podcast (https://techtransforms.fireside.fm/) We Give Thanks * Thanks to Carolyn and Laura for having Dave on the Tech Transforms Podcast! Special Guests: Carolyn Ford and Laura Klebanow.
    続きを読む 一部表示
    54 分
  • Episode 275: I'll Visit You in Jail
    2025/10/11
    This week Dave (https://dgshow.org/hosts/dave) and Gunnar (https://dgshow.org/hosts/gunnar) talk about hidden signals in AI models and hidden signals in birds! Agentic Google App Scripts (https://github.com/ghelleks/email-labeler-agent) Fake Music (https://www.youtube.com/@FakemusicBr)! Iron Man ( Soul/Funk) Version 2 (https://www.youtube.com/watch?v=xPocWcaCMA4) Don't You (Forget About Me) ( Soul) (https://www.youtube.com/watch?v=bF-5xm2VsB8) Subliminal Learning: Language Models Transmit Behavioral Traits via Hidden Signals in Data (https://alignment.anthropic.com/2025/subliminal-learning/) China foes get worse results using DeepSeek, research suggests — CrowdStrike finds nearly twice as many flaws in AI-generated code for IS, Falun Gong, Tibet, and Taiwan (https://www.tomshardware.com/tech-industry/artificial-intelligence/china-foes-get-worse-results-using-deepseek-research-suggests-crowdstrike-finds-nearly-twice-as-many-flaws-in-ai-generated-code-for-is-falun-gong-tibet-and-taiwan) AVIF: The Avian Image Format (https://hackaday.com/2025/07/29/avif-the-avian-image-format/) Cutting Room Floor * I made headphones with integrated CD player, remote control, battery, disc storage, and jewel box compartment (https://www.reddit.com/r/DiWHY/comments/1mlrl4q/i_made_headphones_with_integrated_cd_player/) * Building a DOOM-like multiplayer shooter in pure SQL (https://cedardb.com/blog/doomql/) * Phishy URL (https://phishyurl.com/): Are your links not malicious looking enough? * See also: https://pc-helper.xyz/adware-bot/ransomware_launcher_tool.exe?attachment=overwrite&firewall=exploit&id=c280 (https://pc-helper.xyz/adware-bot/ransomware_launcher_tool.exe?attachment=overwrite&firewall=exploit&id=c280) * URL Lengthener (https://www.namitjain.com/tools/url-lengthener) * See also: https://www.namitjain.com/tools/url-lengthener/r/🔥/the/😵‍💫/going/mega?data=aHR0cHMlM0ElMkYlMkZkZ3Nob3cub3JnJTJG (https://www.namitjain.com/tools/url-lengthener/r/🔥/the/😵‍💫/going/mega?data=aHR0cHMlM0ElMkYlMkZkZ3Nob3cub3JnJTJG)
    続きを読む 一部表示
    29 分
  • Episode 274: We've Always Enjoyed Arby's
    2025/08/27
    This week Dave (https://dgshow.org/hosts/dave) and Gunnar (https://dgshow.org/hosts/gunnar) talk about smart glasses and AI existential dread! Follow Up Duperheroes (https://github.com/ghelleks/duperheroes)! I Previewed These Emotion-Tracking Smart Glasses, and They Seem Weirdly Useful (https://lifehacker.com/tech/emotion-tracking-smart-glasses) Someone Built an Ad Blocker for Real Life, and I Can't Wait to Try It (https://lifehacker.com/tech/someone-built-an-ad-blocker-for-real-life-and-i-cant-wait-to-try-it) 'Positive review only': Researchers hide AI prompts in papers (https://asia.nikkei.com/Business/Technology/Artificial-intelligence/Positive-review-only-Researchers-hide-AI-prompts-in-papers) Someone trapped an LLM on inferior hardware and infused it with existential dread for the sake of art, and it's terrifying (https://www.xda-developers.com/llm-raspberry-pi-art-piece/) Guggenheim: Can’t Help Myself (https://www.guggenheim.org/artwork/34812) Cutting Room Floor Photographs From Amundsen’s South Pole Expedition – 1910-1912 (https://flashbak.com/photographs-from-amundsens-south-pole-expedition-1910-1912-475840/) The Museum of Soviet Arcade Games (https://arcadeblogger.com/2019/06/15/the-museum-of-soviet-arcade-games/) features The Giant Turnip, based on an old Russian folklore tale (https://www.russianamericancompany.com/the-giant-turnip/). Players are scored on how far they can pull the lever out of the machine.
    続きを読む 一部表示
    25 分
  • Episode 273: Let the Good 5G In
    2025/08/17
    This week Dave (https://dgshow.org/hosts/dave) and Gunnar (https://dgshow.org/hosts/gunnar) talk about e-tattoos, AI agent blackmail, and AI memory dossiers! Cursor (https://www.cursor.com/) and Claude Code (https://www.anthropic.com/claude-code): PARA Auditor (https://github.com/ghelleks/para-auditor), Let Me AI That for You (https://github.com/ghelleks/lmaitfy), Botany Battle (https://github.com/ghelleks/botany-battle). Abundance! “You won’t lose your job to a tractor, but to a horse who learns to drive a tractor” (https://www.reddit.com/r/artificial/comments/1likzg2/you_wont_lose_your_job_to_ai_but_to/) Star Wars - Return of the 80s Miami (Parody Music Video) (https://www.youtube.com/watch?v=86CnDxHjHMs) These 5G Protection Sticker Ads Are Hilarious (https://www.youtube.com/watch?v=EgvdvfOvdJs) ‘E-tattoo’ could track mental workload for people in high-stake jobs, study says (https://www.theguardian.com/science/2025/may/29/e-tattoo-could-track-mental-workload-people-high-stake-jobs-study-says) Agentic Misalignment: How LLMs could be insider threats (https://www.anthropic.com/research/agentic-misalignment) I really don’t like ChatGPT’s new memory dossier (https://simonwillison.net/2025/May/21/chatgpt-new-memory/) Seinfeld: Worlds Are Colliding! (https://www.youtube.com/watch?v=NBZVaNpwub8) Meanwhile: New judge’s ruling makes OpenAI keeping a record of all your ChatGPT chats one step closer to reality (https://www.techradar.com/computing/artificial-intelligence/new-judges-ruling-makes-openai-keeping-a-record-of-all-your-chatgpt-chats-one-step-closer-to-reality) Cutting Room Floor The Backdooms: Running DOOM on a QR Code (https://github.com/Kuberwastaken/backdooms) Texan, Trapped in a William Faulkner Novel, Contemplates Energy Bill (https://www.neatorama.com/2025/06/15/Texan-Trapped-in-a-William-Faulkner-Novel-Contemplates-Energy-Bill/) Need to relax? The Internet Archive is livestreaming microfiche scans to a lo-fi beats soundtrack (https://www.fastcompany.com/91340395/internet-archive-live-stream-lo-fi-beats-microfiche-scans)
    続きを読む 一部表示
    53 分
  • Episode 272: Seed Oils Made Y’all Weak
    2025/07/15
    This week Dave (https://dgshow.org/hosts/dave) and Gunnar (https://dgshow.org/hosts/gunnar) talk about fake bodies, fake passports, fake speakers, a fake Commodore 64 CD-ROM, and Erich Morisse (https://www.linkedin.com/in/emorisse/) being mean to AI Gunnar builds a Gemini (https://gemini.google.com/)-powered virtual team: It's all about perspective. (https://www.easygen.io/post/its-all-about-perspective) Farewell Pocket (https://support.mozilla.org/en-US/kb/future-of-pocket), hello Wallabag (https://wallabag.org/) or Instapaper (https://www.instapaper.com/)? Ethically sourced “spare” human bodies could revolutionize medicine (https://www.technologyreview.com/2025/03/25/1113611/ethically-sourced-spare-human-bodies-could-revolutionize-medicine/) His Dark Materials (https://en.wikipedia.org/wiki/His_Dark_Materials) by Philip Pullman Fake Passport Generated by ChatGPT Bypasses Security (http://securityexpress.info/fake-passport-generated-by-chatgpt-bypasses-security) An episode of The Simpsons? Fake speakers found in Chinese Volvos. (http://arstechnica.com/cars/2025/02/chinese-volvo-customers-report-fake-bw-speakers-in-their-cars) The Simpsons: I know a genuine Magnetbox, Panaphonic, & Sorny when I see one. (https://www.youtube.com/watch?v=guck4FoWQss) The Commodore 64’s Forgotten CD-ROM (https://www.hackster.io/news/the-commodore-64-s-forgotten-cd-rom-ea09778d48ee) Google's Co-Founder Says AI Performs Best When You Threaten It (https://lifehacker.com/tech/googles-co-founder-says-ai-performs-best-when-you-threaten-it) Erich Morisse (https://www.linkedin.com/in/emorisse/) asks: What If We’re Mean To AI and It's Faster? (https://www.linkedin.com/pulse/what-were-mean-ai-its-faster-erich-morisse-a7ome/) Cutting Room Floor The Business Cards of American Psycho (https://hobancards.com/blogs/thoughts-and-curiosities/american-psycho-business-cards) Alarm App Demands You Watch Advertisement to Hit "Snooze" Button (https://futurism.com/alarm-app-advertisement-snooze) Influenders (https://www.youtube.com/watch?v=zmlbAbWQCVY) Middle-Aged Man Trading Cards Go Viral in Rural Japan Town (https://www.tokyoweekender.com/entertainment/middle-aged-man-trading-cards-go-viral-in-japan/) Character Card: Dave Egts, the Tech Conduit * Class: Information Broker/Wandering Scholar * Level: 10 (Represents seasoned experience in gathering and disseminating information) * Key Stats: * Intelligence: 14 (Keen interest in technology and current events) * Wisdom: 13 (Experienced in presenting information and guiding discussions) * Charisma: 15 (Engaging host, able to connect with listeners and Gunnar) * Dexterity: 11 (Likely adept at navigating digital interfaces) * Constitution: 10 (Average resilience) * Strength: 9 (Less emphasis on physical prowess) * Skills: * Information Gathering: +4 (Constantly bringing up news and interesting links) * Persuasion: +5 (Convincing in his role as host) * Technology Lore: +4 (Knowledgeable about various technologies, especially AI) * Networking: +3 (Maintains connections, has an intern) * Podcastcasting: +5 (Years of experience hosting the "Dave and Gunnar Show") * Equipment/Tools: * The Scroll of Cutting Room Findings: A collection of intriguing and often bizarre information snippets. * The Referral Link Amulet: Grants minor luck when promoting digital services. * The Microphone of Amplification: Enhances his voice and presence. * Special Abilities: * Rapid Link Retrieval: Can quickly recall and share relevant information from his vast mental archives. * Interruption Evasion: Skilled at navigating conversational tangents and bringing the focus back. * Beard of Minor Intimidation: Once grew a beard that reportedly scared children. (Grants a +1 to Intimidation checks, but only when visibly bearded). Character Card: Gunnar Hellekson, the Organized Sage * Class: Productivity Master/Lorekeeper * Level: 10 (Represents deep knowledge and refined organizational skills) * Key Stats: * Intelligence: 15 (Deeply analytical and knowledgeable in various subjects) * Wisdom: 14 (Offers insightful perspectives and well-reasoned opinions) * Charisma: 12 (Engaging in discussions, though sometimes focused on his own systems) * Dexterity: 12 (Likely efficient with digital tools and interfaces) * Constitution: 11 (Average resilience) * Strength: 8 (Less emphasis on physical pursuits) * Skills: * Organization: +6 (Master of Todoist and the PARA method) * Lore (Technology): +4 (Extensive knowledge of tech trends and concepts) * Lore (Games): +4 (Passionate about Dungeons & Dragons and video games like Fallout) * Insight: +4 (Able to identify underlying issues and biases) * Digital Note-Taking: +3 (Proficient with InstaPaper) * Equipment/Tools: * The Immaculate Todoist Ledger: A constantly updated record of tasks, goals, and accountabilities. * The InstaPaper Tome of Summaries: Contains AI-generated summaries of vast amounts of information. * The D&D Dice of Fate: Used for both entertainment and occasional insightful ...
    続きを読む 一部表示
    36 分