Divya van Mahajan

How to Help People Thrive with AI

ai-daily-brief-podcast
Listen to episode →

Overview

This episode of the AI Daily Brief podcast examines how individuals and organizations can help people genuinely thrive with AI, rather than simply deploying more powerful models. The host (unnamed in the transcript) synthesizes findings from an AI proficiency report, a long-form Atlantic essay by David Brooks, and a real-world case study from Uber to argue that the central challenge of the AI era is not capability but human adoption, motivation, and institutional design.

Source video: URL not provided.

Prerequisites

  • Basic familiarity with AI tools, including large language models (LLMs) and AI agents
  • General understanding of workplace productivity concepts and organizational change management
  • Awareness of current AI adoption trends in enterprise settings
  • Familiarity with terms such as “agentic AI,” “pull requests,” and “vibe coding” is helpful but not essential

Main Points

1. The Gap Between AI Deployment and AI Readiness

  • A recent AI proficiency report found that while 69% of workers reported their organization had taken action on AI agents, only 16% actually use an agentic tool at work.
  • Fewer than 10% of workers can define an AI agent in their own words.
  • Only 30% of employees at organizations with AI agents have received any agentic training.
  • The conclusion: deploying tools without supporting human learning renders the technology largely ineffective.

2. David Brooks’s Atlantic Essay — Cognition, Effort, and AI

  • Brooks argues that in an AI-abundant world, the differentiating factor among people will not be intelligence but their relationship to mental effort.
  • Research cited:
    • ActiveTrack (10,000+ workers): AI adopters experienced more intense work lives — email and messaging time more than doubled, business software use rose 94%, and focused uninterrupted work fell 9%.
    • UC Berkeley Haas: Workers took on previously outsourced tasks because AI made them more accessible.
    • MIT Media Lab: Brain connectivity declined by up to 55% when using ChatGPT versus not using it for similar tasks.
    • Possibility Sciences: Gamma wave activity (a marker of cognitive effort) dropped ~40% when using AI.
  • A phenomenon called “AI brain fry” describes the state of feeling simultaneously productive and mentally drained.

3. Brooks’s Three Archetypes of AI Users

  • Productive Passengers (low need for cognition): Use AI to reduce effort; benefit from productivity gains but risk cognitive atrophy and declining critical thinking skills.
  • Reluctant Optimizers (medium need for cognition): Intend to use AI thoughtfully but succumb to over-reliance under everyday stress; prioritize output over excellence. A GoTo survey found 43% of workers submitted AI-generated content they suspected was error-laden or low quality.
  • Mental Marathoners (high need for cognition): Actively resist AI entropy; seek to use AI to expand their capabilities rather than replace their thinking; value originality and personal agency.

4. The Host’s Critique of Brooks — Using AI for New Things, Not Just Faster Old Things

  • Brooks suggests shaming over-reliance on AI for writing, which the host finds unhelpful and misses the point.
  • The host argues the key distinction is between using AI for tasks you already could do versus using AI to attempt things that were previously impossible.
  • People whose cognitive engagement increases with AI are those who use it as an “opportunity technology” — building agents, tackling unfamiliar domains, and stretching into new capabilities.
  • Mental elasticity, like physical fitness, comes from doing uncomfortable, unfamiliar things — AI raises the ceiling of ambition for what those new things can be.

5. The Role of AI Champions Inside Organizations

  • A Wall Street Journal CIO Journal article identified AI champions — internal “superfans” who receive early access, special training, and executive visibility in exchange for promoting adoption among skeptical colleagues.
  • A law firm example: 60+ champions formalized into a program to track and promote AI adoption.
  • The host’s critique: Champions are often framed as internal PR agents; their real value is not telling colleagues AI is good but showing colleagues what is concretely possible.
  • The host’s 2026 prediction: the emergence of internally deployed “vibe coders” — people who use agentic and coding capabilities to partner with business functions and redesign how work is done, not just accelerate existing workflows by 20%.

6. Uber’s “Agentic Pods” — A Case Study in Organizational AI Integration

  • Uber CTO Praveen Napali reported that 99% of Uber engineers use AI tools and more than 70% of pull requests are attributed to agents.
  • To extend agentic AI beyond engineering (into finance, legal, operations, marketing, HR, etc.), Uber created agentic pods:
    • ~30 AI-proficient engineers paired with domain experts from business functions.
    • Each pod given two weeks:
      • Days 1–2: Shadow the expert, document workflows.
      • Day 3: Prioritize automation opportunities.
      • Days 4–5: Build a working agent alongside the domain expert.
      • Days 6–9: Validate generalizability with additional workers.
      • Day 10: Ship.
  • Results from 16 pods across 16 business functions:
    • Capital allocation across 150 cities: 15 hours → 30 minutes
    • Financial pacing reports: 2 days → 10 minutes
    • Marketing web QA: 2 weeks → 50 minutes
    • Support workflow creation: 9,000 manual workflows → self-service automation
  • Key insight from Napali: The biggest wins came from rethinking entire workflows, not automating individual tasks — eliminating handoffs, unnecessary approvals, and legacy tooling.
  • Key lesson: “The best AI opportunities are rarely visible from the outside. You discover them by sitting next to the people doing the work.”

7. The Host’s Extension of the Uber Model — Reinvesting Productivity Gains

  • The host agrees the two-week pod model surfaces valuable low-hanging fruit quickly.
  • The deeper opportunity, however, lies in what happens after: business people, newly exposed to agentic thinking, begin to reimagine their work at a fundamental level.
  • The real transformation is not the engineer saving 23 hours and 50 minutes on a financial report — it is the business person figuring out what entirely new work to do with that reclaimed time.
  • This “reinvestment of productivity gains” — not the productivity itself — is what will fundamentally change organizations.

8. Disagreement with Brooks on Fixed Motivation

  • Brooks implies that intrinsic motivation (need for cognition) is largely fixed, and only “mental marathoners” will truly thrive.
  • The host disagrees: most institutions have historically asked very little of people, given them discrete tasks for unclear reasons, and defined success narrowly.
  • If AI is deployed with genuine support — stretching, challenging, and incentivizing people — far more latent human potential will emerge than current assumptions suggest.

Key Concepts

  • Agentic AI / AI agents: AI systems that can take sequences of actions autonomously to complete multi-step tasks, rather than simply responding to a single prompt.
  • Agentic readiness: The degree to which an organization’s workforce is trained, equipped, and culturally prepared to work effectively with AI agents.
  • AI brain fry: A colloquial term for the mental state of feeling simultaneously more productive and more cognitively drained as a result of intensive AI use.
  • Need for cognition: A psychological trait describing the degree to which an individual intrinsically enjoys and seeks out effortful thinking; ranges from “cognitive misers” (low) to “mental marathoners” (high).
  • Cognitive miser: An individual who finds effortful thinking unpleasant and will avoid it when possible; Brooks’s term for the low-need-for-cognition archetype.
  • Productive passengers: Brooks’s archetype for low-need-for-cognition workers who use AI to reduce effort; may benefit from productivity gains but risk cognitive atrophy.
  • Reluctant optimizers: Brooks’s archetype for medium-need-for-cognition workers who intend to use AI thoughtfully but default to over-reliance under everyday pressures.
  • Mental marathoners: Brooks’s archetype for high-need-for-cognition workers who actively resist cognitive outsourcing and use AI to expand rather than replace their own thinking.
  • Infinite backlog: The host’s concept describing how AI and agents eliminate natural stopping points in work, creating a perpetual queue of tasks that could theoretically always be addressed.
  • AI champions: Employees designated within an organization to receive advanced AI training and promote adoption among colleagues through demonstration and peer conversation.
  • Internally deployed vibe coders: The host’s term for a predicted role: individuals who use agentic and low-code/no-code AI capabilities to embed within business functions and redesign workflows from the inside.
  • Agentic pods (Uber): Uber’s two-week cross-functional teams pairing AI-proficient engineers with business domain experts to identify, build, and ship agentic automations for non-engineering workflows.
  • Industrialization of detachment: A term coined by educator Chris Seiben describing students’ loss of appreciation for effortful, iterative creative work when AI can produce equivalent outputs instantly.
  • Forward-deployed engineers: A model (referenced in contrast) where AI vendors place engineers inside client organizations to help integrate technology.

Summary

The central argument of this episode is that the proliferation of increasingly powerful AI models is necessary but insufficient: the critical bottleneck is now human adoption, motivation, and institutional support. Drawing on an AI proficiency report, David Brooks’s Atlantic essay on cognition and volition, and Uber’s agentic pods initiative, the host contends that the most important variable determining who thrives with AI is not intelligence but one’s relationship to mental effort — and, crucially, that this relationship is not fixed. While Brooks’s three archetypes (productive passengers, reluctant optimizers, and mental marathoners) illuminate real patterns of behavior, the host challenges Brooks’s implicit pessimism about lower-motivation workers, arguing instead that institutions have historically underestimated and under-challenged people. The host’s prescription is twofold: first, individuals should orient toward using AI to do things they could not do before — stretching into new domains — rather than merely accelerating existing work; second, organizations should invest in embedded, collaborative models like Uber’s agentic pods, which pair technical and domain expertise to redesign workflows from the ground up, and then trust that the productivity gains, once reinvested by newly AI-literate business workers, will unlock entirely new categories of work that were previously impossible to imagine.