Drawing from Harari's book Nexus, Caio and Marina dig into a quiet shift: for the first time, we've built a network that doesn't just distribute information — it interprets it and decides for us. With no fearmongering and no naive optimism, the conversation flips the classic AI question on its head and shows what it actually means for anyone running a business.
In this episode
01 The hook: when information starts deciding
- Open with the staircase: the internet gave us access to information, social media decided what reaches us, and now AI decides what that information means
- Marina asks the listener's question: 'okay, but isn't this just hype from someone who just read a book?' Caio brings the Nexus point in simple terms, without turning it into a book review
- The line that anchors the episode: the danger isn't AI thinking like a human, it's humans not noticing when they're thinking through it
- Set the tone: no robots taking over the world here, just something more subtle happening
02 The shift: from carrying information to interpreting it
- Caio walks through the timeline of networks: the printing press printed, radio broadcast, the internet connected, and AI interprets — a difference in kind, not degree
- Marina asks for a concrete example so it doesn't stay abstract: what changes when a system recommends instead of just storing the data
- The feed example: the social media algorithm has been doing this for years, so generative AI didn't invent the problem, it scales it and gives it a voice
- The turn: AI starts participating in the decision, not just the communication, and we don't even notice
03 The invisible future that's already here
- Drop the humanoid robot image: the real impact is in CRM, hospital triage, bank credit scoring, internal HR systems screening resumes
- Caio gives a business case: when AI decides which lead a salesperson handles first, it's already shaping the commercial result without anyone voting on it
- Marina pushes back: 'but is that bad? Sometimes the machine decides better than we do.' Caio partly agrees and shows where the risk lives
- The invisible layer point: most of these decisions happen without the person even knowing an AI was involved
04 The right question to ask
- Drop the worn-out question 'is AI going to steal jobs' and swap it for 'who designs the systems that influence our choices'
- Caio connects with the leadership audience: if you outsource that entire layer to a vendor, you've outsourced your company's decision-making criteria
- Practical example: two AI-powered CRMs can prioritize customers in opposite ways, and that's a choice made by whoever programmed it, not a neutral truth
- Marina sums up the insight: AI isn't neutral because someone always chose what it's going to optimize for
05 Closing: how to think about this responsibly
- Neither automatic salvation nor apocalypse: AI as a new layer of reality that needs correction, transparency, and humans in the loop
- Caio gives concrete advice: before rolling anything out, ask what the system is optimizing for and who reviews it when it gets things wrong
- Always keep an important decision under human review — not out of fear, but because that's where accountability lives
- Close by returning to the title: AI isn't coming to take over, it's coming to organize reality, and organizing reality is power, so it's worth knowing who holds that power