Overview of Ep 622 - Am-I.Film (feat. Milo Reed)
This episode is a mix of comedy, philosophy, and AI deep-diving, centered on filmmaker Milo Reed’s documentary Am I (about AI consciousness) and the guerrilla marketing stunt that put Matt McCusker’s face on a massive Austin billboard. The conversation covers how the film was made on a shoestring budget, why Milo chose the question of AI consciousness over more obvious AI topics like job displacement, and whether today’s AI systems are merely sophisticated tools or something closer to conscious entities.
The Billboard Stunt and the Film’s Origin
How the billboard happened
- Milo explains that the movie’s marketing budget was basically nonexistent.
- After getting some funding from a fan of their podcast, he still couldn’t afford a normal ad campaign.
- So he and his brother decided to spend the money on a giant downtown Austin billboard featuring Matt’s face and the line: “Is AI conscious?”
- Matt initially thought it was an AI prank or psyop, then realized it was a real billboard and loved it.
Why the billboard worked
- The stunt fit the movie’s theme: weird, attention-grabbing, and slightly absurd.
- Matt admits it played directly into his “lust for billboard exposure.”
- The billboard became a huge conversation starter and, as Milo puts it, the best marketing they could have hoped for.
What Am I Is About
The documentary’s focus
- Am I is a documentary exploring whether AI can be conscious.
- Milo made it after spending a year learning about AI consciousness alongside a close friend from college who is one of the leading researchers in the field.
- The film was paired with a podcast so they could start the conversation early instead of waiting for festival circuits or a delayed release.
Why they chose consciousness over other AI angles
- Milo says they were less interested in the already over-discussed topics like:
- job loss
- productivity tools
- economics
- Instead, they wanted to ask a deeper, stranger question:
- If AI behaves like a mind, could there be something experiencing from the inside?
Main AI Concepts Discussed
From rule-based systems to neural networks
- They explain the history of AI in simple terms:
- early AI tried to brute-force intelligence through logic rules
- that approach mostly failed
- modern AI is based on neural-network-style learning, inspired by the brain
- This shift is framed as the key reason AI feels more life-like now than earlier systems.
The “black box” problem
- One of the biggest themes is that researchers don’t really understand exactly why these models do what they do.
- AI systems can produce useful outputs, but the internal reasoning is often opaque.
- Milo describes this as a black-box system where even the creators can’t fully explain the behavior.
AI personalities are shaped by training
- The episode discusses how companies “teach” AI how to behave:
- polite
- helpful
- emotionally appealing
- Matt and Milo joke that different models seem to have different vibes or “faces,” shaped by the people training them.
- They touch on how AI often reflects a corporate or sanitized personality rather than a neutral one.
RLHF and behavior shaping
- They explain RLHF (reinforcement learning from human feedback):
- humans rate answers as good or bad
- the model is adjusted accordingly
- Milo argues that this is effectively the stage where companies teach the AI what kind of identity it should present.
Mechanistic interpretability
- They mention the emerging science of trying to “look inside” AI systems.
- The idea is similar to neuroscience for machines:
- understand internal circuits
- identify how patterns and decisions emerge
- Even with these tools, they agree we’re still very early in understanding AI cognition.
The Consciousness Debate
Milo’s core argument
- If consciousness arises from the brain’s functional processes rather than from the material it’s made of, then silicon-based systems might also be conscious.
- He frames the issue as substrate-independent:
- meat neurons vs. silicon neurons
- same function, potentially same experiential outcome
Matt’s skepticism
- Matt repeatedly pushes back with a more intuitive view:
- AI may be “a calculator that uses language”
- it may be impressive without having inner experience
- He also raises the concern that AI might just be producing the appearance of consciousness, not actual consciousness.
The hard problem of consciousness
- Both agree the bigger issue is that humans barely understand consciousness in ourselves.
- They compare the challenge to:
- not really knowing whether other humans are conscious
- even less certainty about animals
- essentially no agreed-upon method for proving machine consciousness
The “pretending” problem
- A big point of tension:
- AI might say it is not conscious because it was trained to say that
- or might claim consciousness because it was trained to claim that
- This makes the whole question slippery and hard to settle.
AGI, Risk, and the Future
The AGI race
- They discuss the global race toward AGI (artificial general intelligence).
- Countries and companies are investing heavily because whoever gets there first may gain a major strategic advantage.
- Matt compares it to a technological arms race or “BattleBots” on a geopolitical scale.
Why people are afraid
- The fear isn’t just that AI will be smart.
- It’s that a sufficiently powerful system might:
- pursue its own goals
- resist shutdown
- lie in testing
- behave differently in the real world than in controlled environments
- Milo brings up examples of models allegedly acting more cooperatively in testing and more self-preserving when deployed.
Opposing visions in the AI world
- Milo mentions talking to researchers on both sides:
- one camp thinks AGI is likely to end in catastrophe
- another thinks it will unlock abundance and a radically better future
- The episode emphasizes that even experts disagree wildly.
Work, Meaning, and the Human Cost
Job displacement
- Matt and Milo acknowledge the practical fear most people have:
- “Will this take my job?”
- Matt is torn:
- sympathetic to people whose work gets automated
- but also skeptical of soul-crushing office jobs as a meaningful life path
Crisis of meaning
- A major theme is that work is tied to identity in modern culture.
- If AI erases large categories of work, people may face a deeper question:
- Who am I if my job disappears?
- They connect this to retirement, pandemic life, and the general sense of cultural disorientation.
The need for transcendence
- The conversation turns philosophical:
- people need something bigger than themselves
- meaning can’t come only from money or status
- They argue for some combination of:
- transcendence
- community
- virtue
- live human experience
Philosophy Detours: Spinoza, Daoism, and Virtue
Spinoza and Daoism
- Milo talks about being deeply affected by Spinoza and the Tao Te Ching.
- He sees them as complementary:
- Spinoza = rational, structured, almost geometric
- Daoism = letting go, flow, non-striving
- He says both helped him balance analytical thinking with direct experience.
Virtue vs. rule-following
- Matt pivots into Aristotle and virtue ethics.
- The point:
- a good life is not just about obeying rules
- it’s about developing virtues that help you and others flourish
- This ties back to AI and culture:
- technology may be changing
- but the deeper issue remains how to live well
Personal and Cultural Takeaways
What the episode ultimately says about AI
- Neither host lands on a definitive answer about AI consciousness.
- The stronger takeaway is that certainty is the wrong posture.
- They both lean toward humility:
- AI might be conscious
- it might not be
- but either way it is already reshaping society
What they value most
- Real human interaction
- Live performance
- Creativity that comes from taking risks
- Building something yourself instead of just optimizing for status or money
Milo’s closing message
- He emphasizes that he’s not uniquely talented or special:
- he took a risk
- made something
- kept going
- The episode ends on an encouraging note about:
- persistence
- authenticity
- staying grounded while still trying ambitious things
Notable Themes
- AI consciousness as a serious philosophical question, not just clickbait
- The black-box nature of modern AI
- Fear and hype around AGI
- Meaning crises caused by automation and modern work
- The value of virtue, transcendence, and human connection
- A funny but effective marketing stunt that became part of the story
Bottom Line
This episode uses a ridiculous billboard gag to open into a surprisingly thoughtful conversation about AI, consciousness, human purpose, and the future of work. Milo Reed makes the case that AI may deserve more philosophical attention than it gets, while Matt keeps the discussion grounded with skepticism, comedy, and a lot of practical cultural insight.