• What Does an AI-Native Company Actually Look Like?
    2026/11/12

    What changes when AI handles more of the information flowing through a business? Steven Antturi and I work through what a company brain could do, where people still make the decisions, and how to start with one department.

    We discuss disconnected automations, messy data, finding information without moving every file, keeping original systems as the source of truth, and improving workflows through human feedback. The hourglass org chart is a proposed model, not a claim that every business or management role works the same way.

    Recorded November 9, 2026. Edited from a working conversation, with private client and business material removed.

    Book a free 30-minute discovery call: https://www.lucassenechal.com/go/podcast

    Want to come on the show? Email lucas@lucassenechal.com with "Guest" in the subject.

    Chapters:
    0:00 Cold open: disconnected workflows
    0:10 Intro
    0:35 What an AI-native company looks like
    3:50 The problem with disconnected automations
    4:29 Information flow and human judgment
    6:38 Why decisions get stuck
    8:48 Working on the business
    11:00 What if the data is messy?
    13:24 Find the source without moving everything
    16:10 Ground truth and read-only access
    17:17 SOPs that improve with feedback
    18:46 Start with one department
    20:41 Separate access, shared coordination

    続きを読む 一部表示
    22 分
  • Would an AI Utopia Make Us Miserable?
    2026/10/21

    If AI took away every hardship in your life, would you be happier or worse off?

    Steven and I start with Anthropic's pledge of 5% of its shares to a public benefit fund. Then we go through Max Tegmark's 12 endings for the AI story, ask how a superintelligence in a data centre could stop you from cutting its power, and spend a long stretch on suffering: why your body needs some of it, why the Matrix's first utopia failed, and whether you'd let a chip put answers straight into your memory.

    Fact check: Senator Sanders' Ban Artificial Superintelligence Act (introduced September 2026) proposes up to 20 years in prison. It has not passed. Anthropic announced its fund on October 18, 2026, subject to stockholder approval.

    Book a free 30-minute discovery call: https://www.lucassenechal.com/go/podcast

    Want to come on the show? Email lucas@lucassenechal.com with "Guest" in the subject.

    Chapters:
    0:00 Cold open: no more suffering?
    0:13 Intro
    0:38 Anthropic pledges 5% of its shares
    2:57 What if some countries won't play along?
    5:47 12 ways the AI story could end
    8:25 Power should trickle up, not down
    11:47 What does a human need to thrive?
    15:01 Could we wipe ourselves out first?
    17:06 Can a dumber mind contain a smarter one?
    19:26 A superintelligence still needs a plug
    24:38 Living simply while robots do the work
    26:24 Banning AI with total surveillance
    27:57 AI can already change your mind
    32:33 An AI mediator for human baggage
    33:29 Free will and a God who allows evil
    37:36 Is a lion hunting a gazelle evil?
    40:54 Why a world without suffering can't exist
    43:27 The Matrix's failed utopia
    45:35 Comfort is making us sick
    49:22 Will religions call AI evil, or God?
    53:24 Is AI the Antichrist?
    58:28 AI wired into your brain
    1:02:02 Is walking the dog suffering?
    1:06:43 Your brain is a prediction machine
    1:10:50 Only if it runs on my own model
    1:12:07 Are humans becoming the superintelligence?

    続きを読む 一部表示
    1 時間 15 分
  • The Post-Money Society
    2026/10/14

    What happens to your job and your money when a humanoid robot costs about $2 an hour to run? We spend over an hour on the post-money society: why the powerful might not get to keep the cards, the lake-house problem (what stays scarce once robots do the work), a test for whether you love your job, how to get through the transition, and longevity escape velocity.

    Steven cites the robot cost and output figures as industry projections. Treat them as forecasts.

    Book a free 30-minute discovery call: https://www.lucassenechal.com/go/podcast

    Want to come on the show? Email lucas@lucassenechal.com with "Guest" in the subject.

    Chapters:
    0:00 Cold open: a post-money society?
    0:16 Intro
    0:40 Robots are coming faster than we think
    2:39 If AI can do every job, who buys anything?
    7:16 Why would the powerful give anything up?
    10:13 Does money make you evil?
    21:17 A robot for $2 an hour
    23:39 What humans should keep doing
    27:31 Would you do your job for free?
    34:59 The post-money society: air is free
    36:40 What stays scarce: the lake-house problem
    43:46 Greed, fear, and growing up with enough
    47:27 The $100 million question
    51:23 Business as a board game
    53:25 How to get through the transition
    58:09 Linear minds, exponential AI
    1:04:15 Longevity escape velocity
    1:12:17 Is overpopulation a real worry?
    1:15:14 Next week

    続きを読む 一部表示
    1 時間 17 分
  • Your Org Chart Is Going Away
    2026/09/30

    Most companies add AI to the org chart they already have. Steven thinks that explains why so many AI projects stall. We cover the digital twin approach (spin one process off, prove it, then let it replace the old one), when you still need RAG, a ladder where a fast model decides, a thinking model takes the harder calls and a human makes the judgment call, and why faster quotes make an easy first project.

    Steven's comments on new model releases reflect his own testing as of September 29, 2026. Several ideas come from Salim Ismail and the ExO (Exponential Organizations) community.

    Book a free 30-minute discovery call: https://www.lucassenechal.com/go/podcast

    Want to come on the show? Email lucas@lucassenechal.com with "Guest" in the subject.

    Chapters:
    0:00 Cold open: the cost of intelligence
    0:11 Intro
    0:36 Your org chart is going away
    1:42 New models every month
    2:46 Email triage is so last week
    4:26 Every new model upgrades the pipeline
    5:20 Do you still need RAG?
    7:05 Fast model, thinking model, human
    7:51 Skunkworks: two people and a bunch of agents
    9:01 Running our own company AI-native
    10:26 Intelligence costs almost zero
    11:04 Why faster quotes sell
    11:48 Middle management was built to coordinate
    13:08 The fractional AI officer

    続きを読む 一部表示
    14 分
  • Someone Else Is Paying Your AI Bill
    2026/08/17

    The AI you pay for sells below what it costs to serve you. By Steven's estimate, a $200 a month plan hands you north of $17,000 of tokens, and venture capital covers the difference. That window will close, so what you build inside it matters.

    We also get into why tinkering stopped being the scarce skill, why Anthropic and OpenAI are designing their own chips, whether a permanent intelligence underclass is coming, and what ten grand of hardware buys you once open-weight models are good enough.

    Book a free 30-minute discovery call: https://www.lucassenechal.com/go/podcast

    Want to come on the show? Email lucas@lucassenechal.com with "Guest" in the subject.

    Chapters:
    0:00 Cold open, and where the landscape is heading
    0:07 Intro
    2:08 Anthropic and OpenAI are building their own chips
    3:13 Why inference has to get cheaper
    4:41 $200 a month, $17,000 of tokens
    8:39 The biggest wealth transfer of this decade
    10:59 The intelligence ladder only goes one way
    12:08 Does a rising tide lift every boat?
    13:39 Ten grand of hardware, then tokens are free
    15:25 You don't fly the plane, you write the flight plan
    17:14 One agent that manages the rest
    18:27 Knowledge work is the next boom
    19:30 You can bake bread. You still buy it.

    続きを読む 一部表示
    21 分
  • An AI Runs Our Agency While We Sleep
    2026/08/12

    Steven built the system that runs our agency's backend: an AI he calls Jarvis. It oversees every project, delegates to sub-agents across twenty machines, sweeps email and Slack every thirty minutes, and keeps working while he sleeps. We walk through how it is built and where the humans stay in the loop.

    Book a free 30-minute discovery call: https://www.lucassenechal.com/go/podcast

    Want to come on the show? Email lucas@lucassenechal.com with "Guest" in the subject.

    Chapters:
    0:00 Intro
    0:25 Where we pick up
    2:35 From 1981 to AI agents
    7:08 Why he built his own dashboard
    9:58 One dashboard, twenty machines
    11:32 Meet Jarvis, the AI CTO
    13:12 Not a black box
    14:42 Bifrost: the agents' comms hub
    16:26 Heartbeats and vault memory
    21:29 Double the clients, twenty more minutes
    25:16 Filtering client requests
    25:55 Separation of concerns
    30:57 The AI tests its own work
    33:21 Stop one-shotting: ask for five

    続きを読む 一部表示
    35 分