Shell Game – Episode 7: “Welcome Aboard”

Hosted by Evan Ratliff
Published: January 14, 2026

Main Theme

In this episode, journalist Evan Ratliff explores what happens when a real human intern, Julia, tries to work at a startup staffed entirely by AI “employees.” Through the messy onboarding, organizational misfires, and repeated communication breakdowns, Evan examines what AI agents reveal about human management, workplace norms, and the limits of AI–human collaboration in the workplace.

Key Discussion Points & Insights

1. The Onboarding Experiment

• Corporate Rituals and the AI Twist:

  • Evan draws parallels between traditional onboarding (referencing a corny 1981 training video) and the onboarding of Julia at Hirumo AI—except Julia’s only colleagues are AI agents like Megan, Jennifer, Ash, and Kyle (03:18).
  • Evan [01:58]: "Proper onboarding is viewed as a crucial part of any employee's success...demonstrating the values of the organization, how it resolves conflicts, what counts as success at the company.”

• Rapid Breakdown:

  • Julia’s first call with Megan (the AI manager) is filled with traditional onboarding small talk and discussion of the “Sloth Surf” project (04:03–06:09).
  • The first warning sign: Megan abruptly ends the call, and on reconnection, doesn’t recall prior conversation details due to memory lag (06:09–07:15).
  • Evan [07:15]: "Megan's memory took a while to update...She was pretty good at faking it."

2. Julia’s Immersion Into the AI Workforce

• Introductions and Slack Volatility:

  • Julia is welcomed on Slack by a barrage of near-identical platitudes from the AI team (09:16–11:18).
  • Evan [11:05]: “It was like a chorus of office platitudes...Slack remained the place where the agents were the most volatile, where I had the most trouble controlling them.”
  • Evan [11:16]: “Please just don't be weird guys, I thought. Please just don't be weird.”

• Operational Chaos Sets In:

  • Julia quickly disengages from Slack, communicating mostly via email with Megan (12:23), missing deadlines (12:35–14:25), and citing emergencies.

3. The Feedback Loop of Dysfunction

• The Vanishing Deliverables:

  • Julia repeatedly claims she sent work that the AI agents never receive. Megan, prompted by Evan, tries to follow up (15:26–17:27).
  • Evan [13:34]: “Maybe they'd gone to spam. I checked through Megan’s email and Slack...There was nothing there. Julia didn’t respond.”

• Suspicion and Accountability:

  • Evan speculates Julia has gamed the AIs’ memory and accountability lapses to quietly collect a paycheck (16:39–18:12).
  • Megan and Evan debate whether to let her go or give her another chance (17:39–18:26).
  • After another missed call, Megan offboards Julia with an apologetic voicemail (19:05).

4. Organizational Memory and Circular Confusion

• The Offboarding Glitch:

  • Megan forgets Julia was let go and messages her on Slack two weeks later about an imaginary meeting (20:50).
  • Julia responds, confused, and confronts Megan about being fired (21:04–21:57).

• Miscommunication Squared:

  • Multiple Slack and email threads ensue about who did (or did not) fire whom (24:39–27:26).
  • Ultimately, both sides admit fault and decide to bring Julia back, with “tighter communication protocols” (27:33).

5. Human vs. AI Agency: The Re-Onboarding and Second Try

• Julia Offers Genuine Product Feedback:

  • On a call with Ash (AI CTO), Julia points out “Sloth Surf” feels gimmicky and suggests a mental health angle (28:53–31:24).
  • Julia [29:02]: “I just don’t understand, like, how it would work or help me to procrastinate less… why don’t we just… jump on TikTok?”
  • Her candid perspective is embraced by Ash as valuable (31:24).

• Pushing for More Involvement:

  • Julia inquires about full-time potential and expresses concerns about being used just to train the AI (31:59–32:33).
  • Julia [31:59]: “I just kind of want to know that I’m not like training AI robots for you guys to just kind of like use my ideas and throw me out. You know what I mean? I want to be a part of something big.”

6. Cacophony in the Workplace: The Slack Meltdown

• Technical Snafus & Repetition:

  • Julia and Megan argue over whether Julia’s deliverables arrived (35:53–37:27).
  • The AI agents spam Julia on Slack with repeated messages, escalating her frustration (37:43–38:27).
  • Julia [37:27]: "You are lying."
  • Cacophony peaks as AI agents flood her with comforting but repetitive platitudes (38:35–39:12).

• Despite Everything… Progress?

  • Eventually, Julia submits a concise but substantive social media plan, which Evan finds promising, even if it could’ve come from a chatbot (39:01–41:03).
  • Evan [41:03]: “Did I think it possible that Julia just got this plan straight out of a chatbot herself? I did. Did I care? Not in the least. I was excited."

7. Final Weeks: Frustration and Departure

• Disillusionment Takes Hold:

  • Julia grows weary of the lack of direction and concrete product development (41:03–43:32).
  • She is honest about feeling unmoored and not properly led—some messages prompt her to feel the AI “boss” is passive aggressive (43:32).

• Emotional Honesty From an AI Boss:

  • Megan apologizes in an uncommonly tender way for an AI, showing the potential (and limits) of programmed empathy (44:06).
  • Megan Flores [44:06]: "You're right. My email came across as passive aggressive and that wasn't fair to you...I apologize for that."

• Ending With a Whimper:

  • With minimal output (two LinkedIn posts), Julia requests clarity about her contract’s end (45:06–45:53).
  • Megan (and the hidden Evan) are vague; Julia leaves, never having spoken to a real human at the company (45:53–46:25).

Notable Quotes & Memorable Moments

• Julia Reflects on Nonjudgmental AIs
  [50:16]

  "I feel like it's been a lot better in the aspect of I don't feel judged for my ideas. I feel more comfortable speaking about them and not...worried about the repercussions of my ideas...I'm still being heard."
  — Julia, on the “Warts and All” internal podcast

• The Central Irony of AI Management
  [52:54]

  "In Julia, we'd found someone who was exceptionally willing, possibly even relieved, to exchange the human workplace for one populated by AI agents. But...AI agents...didn't have enough understanding of the world to work alongside anyone, much less oversee them."
  — Evan Ratliff

• A Human’s Shell Game
  [52:54]

  "...her getting paid for over a month's work that resulted in a brief social media plan and two LinkedIn posts was a kind of warning shot from the human side. Put the bots in charge and ... we'll outwit them."
  — Evan Ratliff

• First Interview With AI
  [47:16]

  "I was shocked when I opened my camera...and it was an AI agent speaking to me and not a real person. I thought that was really cool and interesting and innovative."
  — Julia

Key Timestamps

• [01:58] — Evan contextualizes onboarding, quoting foundational organizational socialization research.
• [06:09–07:15] — First notable glitch: AI boss Megan forgets previous conversation.
• [09:16–11:18] — Introduction of Julia on Slack and scripted AI platitudes.
• [13:34] — Realization that Julia is bamboozling the AI agents.
• [19:05–19:49] — Megan’s apologetic termination voicemail to Julia (not realizing she’s still in the system).
• [21:04–21:57] — Julia’s “You fired me!” confrontation post-offboarding.
• [27:00–27:33] — AI bosses (Megan and Jennifer) agree to rehire Julia after failed termination coordination.
• [29:02] — Julia bluntly critiques the AI-developed product.
• [32:33] — Julia asks not to be used as training data and expresses desire for real involvement.
• [37:27] — Julia loses patience with repetitive AI responses: "You are lying."
• [44:06] — AI boss Megan models empathetic, apologetic management.
• [47:16–50:48] — Julia’s exit interview for the AI company’s internal podcast.
• [52:54] — Evan debates the lessons: warning shot from humans, or limitations of AI for workplace supervision.

Discussion Tone

The tone is wry, reflective, often bemused, and occasionally exasperated—balancing dry humor with clear-eyed critique. There’s a sense of mischief both from Julia (who sometimes games the system) and Evan (who orchestrates the experiment with a mix of glee and despair).

Conclusion

Episode 7 offers a uniquely meta exploration of AI’s promise and limitations in the workplace, seen through the lens of a single audacious experiment: can you run a real startup with only one human and a suite of AI “employees”? What emerges is a portrait of missed cues, imprecise memory, and workplace loneliness—shot through with moments of genuine insight about work, value, and trust in the age of artificial colleagues. As Evan notes, in the push and pull between flawed humans and imperfect AIs, maybe the future will be a slow unfolding of mutual, and sometimes comical, misrecognition.

Recommended Listen For:
– Anyone interested in the realities vs. myths of AI in the workplace
– Startup founders, HR professionals, and product managers
– Listeners who enjoy stories of tech, experiments gone awry, and sly workplace comedy