Deep Dives by Artificially Intimidating

Deep Dives by Artificially Intimidating

Download on the App Store

Deep Dives by Artificially Intimidating episodes

  • Every New Tool Was Going to Kill Art, with John Hardin
    This is a free preview of a paid episode. To hear more, visit artificiallyintimidating.com

    Every new creative tool was supposed to kill real art: the printing press, the piano, the microphone, the tape machine, the 808, and now AI. None of them did.

    Artificially Intimidating is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

    John A.C. Hardin is a Senior Fellow at the Abundance Institute, a historian, a working musician, and the writer of the Creative Frontiers newsletter. A few weeks ago I watched him run a panel on AI and music in a basement on the Lower East Side. He opened with a story about Bing Crosby, Les Paul and one tape machine, and I've been thinking about it ever since. So I asked him on the show to keep going.

    We get into why Bing Crosby funded tape so he could play more golf, a monk who was sure the printing press would destroy culture, the drum machine that flopped its way into defining hip hop, why the player piano created more piano teachers, Barry Manilow's synthesizers and the Broadway "walkers," karaoke, the moonwalk, sampling as theft or tribute, Bad Bunny and the dembow lawsuit, and what all of it suggests about AI, ownership and creative work. John also doesn't let me off easy: the disruption is real, and people do lose their jobs.

    Chapters

    0:00 Intro: every tool was going to kill art

    1:14 Meet John Hardin

    3:34 Bing Crosby, Les Paul and the tape machine

    7:55 The printing press panic, and the script that keeps repeating

    9:32 Pianos, valves and metronomes: "the end of music"

    12:35 Photography, Kodak and a camera for everyone

    14:46 How AI gave a photographer new superpowers

    19:56 How John uses AI: writing and a Suno theme song

    23:03 Who owns an AI-assisted song?

    25:11 The 808: a commercial flop that defined hip hop

    27:35 Parody songs, and is it worth it?

    28:56 Why AI labels are "fairly useless"

    30:43 AI and the business side of being an artist

    32:57 The disruption is real: Barry Manilow and the "walkers"

    34:48 Net positive: karaoke and the jobs nobody predicted

    36:07 "This is so dumb": the first internet video I ever saw

    41:33 From civil engineering to storytelling

    43:40 Is copyright a barrier or an incentive?

    45:14 The player piano made more piano teachers

    46:52 Training data and owning what you learn

    49:04 McCartney, Michael Jackson and the Beatles catalog

    49:40 Deepfakes, the NO FAKES Act and owning your face and voice

    51:08 Suno as an instrument

    52:34 Herbie Flowers, A Tribe Called Quest and who gets paid

    53:35 Anthropic, the books and the $3,000 settlement

    54:20 How copyright grew out of the written word

    55:50 What if you could copyright a style?

    56:37 One legal decision's butterfly effect

    58:12 The moonwalk, and dance moves nobody owns

    59:08 Jokes, comedians and borrowed bits

    59:59 Recommendations: Arewa, McLuhan and The Dispatch

    1:02:23 Les Paul, "Lover" and the multitrack leap

    1:05:00 John Mayer and an old nightlife photo archive

    1:05:53 Sampling: theft or tribute?

    1:07:10 Who owns AI-made work?

    1:08:07 Bad Bunny and the dembow lawsuit

    1:09:18 Where to find John

    About John

    John A.C. Hardin, Ph.D. is a Senior Fellow at the Abundance Institute, where he leads Creative Frontiers, an initiative focused on artists, technology, and the future of creative work in the age of AI. A historian, musician, and engineer by training, he studies how new tools reshape who gets to create. His writing has appeared in The Wall Street Journal and The New York Times, and he is the author of Church Advertising, Public Relations and Marketing in Twentieth-Century America: Retailing Religion. He holds a PhD in U.S. history and lives in the Carolinas, pickin' and singin' whenever possible.

    * Abundance Institute

    * Creative Frontiers on Substack

    * Instagram

    * LinkedIn

    * X

    John's writing we talk about

    * The Golfer and The Grinder: Bing Crosby, Les Paul, and a box from Nazi Germany

    * The Drum Machine that Sounded Fake, then Changed Music (Roland TR-808)

    * The Empty Orchestra (the karaoke machine)

    * Bad Bunny and the Billion-Dollar Drumbeat (Borrowed Genius, Part 1)

    * Michael Jackson Didn't Invent the Moonwalk (Borrowed Genius, Part 2)

    * The 1925 New Tech That Let a Legend Invent a New Sound (Duke Ellington and the microphone)

    Also mentioned

    * My write-up of the panel where I met John, Everyone's Job Looks Easy From the Outside

    * Olufunmilayo Arewa, "From J.C. Bach to Hip Hop: Musical Borrowing, Copyright and Cultural Context"

    * Everyday Abundance with Virginia Postrel and Charles Mann (John wrote the theme)

    * The Dispatch Podcast

    * Les Paul & Mary Ford, "Lover" and "How High the Moon"

    * John Mayer, Where the Light Is: Live in Los Angeles

    Correction

    At 1:05:26 I say a Washington Post article showed training data "from where OpenAI got it." The Post's 2023 analysis actually covered Google's C4 dataset, which was used to train Google's T5 and Meta's LLaMA. OpenAI does not disclose its training data. Washington Post analysis.

    Support the show

    Viktor, $50 off your first month

    Full transcript

    Lightly cleaned (filler words removed). Timecodes match the episode.

    1 hr 11 min
  • The Boring Layer That Decides If Your AI Survives

    The first human-to-human interview on Artificially Intimidating. I sit down with Bruno Perez, co-founder and CEO of Manifest — the open-source LLM router that keeps your AI agents running when a provider goes down.

    --

    We get into why one in twenty AI calls fails in production, the self-healing "Autofix" layer Manifest just shipped, how two self-taught founders pivoted the week Claude Code launched, and why nobody's loyal to a model anymore.

    Guest: Bruno Perez, co-founder & CEO of Manifest — https://manifest.build

    In this episode:

    From a 15-year WordPress agency to an open-source AI router

    What Manifest does: one endpoint, your subscriptions + API keys, fallbacks and spend limits

    "Five percent of all LLM calls in production are an error" — and Autofix, the insurance layer

    Paid plans, and why good products deserve to be profitable

    Model loyalty is a fiction, agents spamming agents, and building only the slice of a tool you use

    Links:Manifest — https://manifest.buildAutofix (early access) — https://manifest.build/autofix/Paid plans — https://manifest.build/blog/introducing-paid-plans/GitHub — https://github.com/mnfst/manifestFree LLM API list — https://github.com/mnfst/awesome-free-llm-apisHermes — https://hermes-agent.nousresearch.com/Paperclip — https://paperclip.ing/OpenClaw — https://openclaw.ai/OpenCode — https://opencode.aiNanobot — https://nanobot.wiki/home/Craft — https://agents.craft.do/Paseo — https://paseo.sh · Ghostty — https://ghostty.org · Orgo — https://orgo.ai · Chatwoot — https://chatwoot.com

    Full written deep dive, diagrams, and my own fallback setup (paid): More at https://artificiallyintimidating.com



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit artificiallyintimidating.com/subscribe
    1 hr 8 min

About Deep Dives by Artificially Intimidating

From the publisher's feed

Discover AI's Brighter Side