Revisiting “What is Philosophy?

”
The instructor moved into the intellectual portion of the class by revisiting the central question
from Session 1: “What is philosophy?” This was more than a warm-up—it was a deliberate
attempt to make students re-engage with the foundational question and test how much had been
internalized.

Philosophy, the instructor stressed, is not merely about collecting facts or memorizing
information. It is about active participation, conceptual clarity, and the constant re-examination
of ideas. Unlike scientific methods that aim to verify factual claims, philosophical inquiry focuses
on the meanings and implications behind those claims.

When asked to recall last week’s discussion, the class fell silent. The instructor used this as a
teachable moment: passive listening is incompatible with philosophy. To “do philosophy” is to
question, to test assumptions, and to engage—even if that means risking being wrong. Silence, in
this context, is not neutral; it stalls the dialectical process.

This framing set the tone for the rest of the session. If philosophy is about “thinking about
thinking,” then today’s topics—humans versus machines, thinking as symbol manipulation, and
the nature of emotions and consciousness—would require the class to look past easy answers.
The aim was to push beyond “first thoughts” and into sustained, critical reflection.

In essence, this short exchange reminded everyone that philosophy is not a spectator sport. It’s
an active, sometimes uncomfortable, process that rewards the willingness to probe deeper.

Humans vs. Machines – Initial Differences vs

The first substantive exercise asked students to name differences between humans and machines.
The answers came quickly:

• Humans are emotional; machines are not.

• Humans are creative; machines just follow programmed patterns.
• Humans are unpredictable; machines are predictable.
• Machines can process faster than humans.
• Humans can assign work to machines, but not the other way around.

The instructor treated these answers as hypotheses, not conclusions. He repeatedly challenged
students with “Why not?”—forcing them to defend or reconsider their claims. For example, if a
machine could theoretically assign a task to a human (e.g., through automated scheduling or
instructions), why should that be impossible?

The conversation revealed that many “obvious” differences are actually unexamined
assumptions. What counts as creativity? Can unpredictability be engineered? Are emotions
necessary for certain types of decision-making? By probing these questions, the instructor shifted
the discussion from “listing differences” to “defining terms.”

This moment illustrated a key philosophical skill: questioning the certainty of our own categories.
It’s easy to assume that humans and machines are separated by a clear line, but that line starts to
blur when we scrutinize the underlying definitions.

1
Thinking as Symbol Manipulation
The instructor introduced a functional definition of thinking: the manipulation of symbols
according to rules, producing propositions. In this view, thinking is not mystical—it’s a
structured process.

Humans, for example, combine sensory inputs, stored memories, and concepts into
propositions. The instructor gave a concrete case: recalling a trip to New York might involve
mental images of the Brooklyn Bridge, Times Square, Central Park, and museums. Those images
are assembled into statements like “I went to Central Park at 3 p.m.”

Then came the twist: the instructor revealed that the class notes some students had received were
not written by him directly but generated by GPT from his lecture recordings. This meant a
machine had also taken inputs (audio transcript), applied rules (language processing), and
produced structured propositions.

If thinking is defined as symbol manipulation, the gap between human and machine thinking
narrows dramatically. Machines can perform this kind of processing—sometimes more
efficiently and without fatigue. The philosophical challenge is whether complexity, self-
awareness, or other factors create a meaningful distinction between human and machine
thought.

This reframing unsettled the easy assumption that thinking is uniquely human, setting the stage
for debates on consciousness and emotions.

Symbolic Thinking Beyond Humans

The discussion moved beyond humans and machines to consider other forms of symbolic
processing. The instructor argued that symbolic thinking is not exclusive to humans. Even
simple organisms, such as bacteria, operate on basic symbolic “if-then” structures—for instance,
“detect toxin → move away.”

Students expressed surprise at this idea, revealing how our intuitions about thinking are often
anthropocentric. The instructor used this reaction to emphasize the importance of functional
definitions: if thinking is defined as symbol manipulation, then the degree of complexity may
vary, but the basic mechanism can exist in very simple systems.

This example also extended the discussion toward machines: if bacteria “think” in a basic sense,
then perhaps the threshold for calling something a “thinker” is lower than we usually assume. It
challenges the sharp divide between human cognition and non-human processes, suggesting that
thinking might exist on a continuum—from bacteria to AI—rather than in a single leap from
nothing to full human reasoning.

Defining a Machine
When asked to define “machine,” most students instinctively answered: “something that reduces
human effort.” The instructor smiled but quickly pointed out that this is a common-sense
definition, not an engineering or philosophical one. Such definitions are useful in everyday
conversation but lack the precision needed for rigorous debate.

2
He proposed a more universal and functional definition: a machine is a system that takes an
input, processes it, and produces an output. This definition is intentionally broad. It could
include a steam engine, a coffee maker, a computer, or even a biological organ like the heart.

To make it concrete, he gave the example of a vending machine: insert money (input), select an
item (processing), and receive the snack (output). This model strips away the “human effort”
angle and focuses on the essential structural features of machines.

By shifting from a purpose-based definition (“reduces human effort”) to a process-based one

(“input → process → output”), the instructor made it clear why precise definitions are so
important in philosophy. Without them, discussions about whether machines can think, feel, or
reproduce risk becoming muddled by inconsistent usage of terms.

The discussion also hinted at a bigger philosophical point: if machines are defined purely by
functional criteria, then humans—who also take inputs (sensory data), process them (neural
activity), and produce outputs (speech, action)—might themselves fit the definition of a
“machine.” This provocative implication blurred the boundaries between human and artificial
systems.

Defining Emotions
Before tackling whether machines could have emotions, the instructor insisted that the class
agree on what “emotion” means. Several definitions were offered:

• A response to stimuli (e.g., feeling scared when hearing a loud noise).

• A subjective state tied to circumstances (e.g., feeling joy at a celebration).
• A form of compressed thinking that quickly evaluates a situation.

The instructor emphasized that emotions are internal states whose effects can produce actions,
but they are not the actions themselves. For example, anger might cause a person to shout, but
shouting is merely an expression of the anger, not the emotion itself. Similarly, hugging someone
might be the result of love, but the hug is not the love.

He stressed that without a precise definition, we risk making vague or contradictory claims about
machine emotions. If emotions are defined solely as “responses to stimuli,” machines already
qualify. If they require “subjective feeling,” then the debate hinges on whether subjective states
can exist in non-biological systems.

This definitional work mirrored a core philosophical method: before debating a possibility,
one must clarify the concept under discussion. Otherwise, disagreements may simply be
about language, not reality.

Can Machines Have Emotions?

With provisional definitions in place, the instructor opened the floor to debate. The class split
into two camps:

Yes, machines can have emotions

• If emotions are simply responses to stimuli, machines already replicate them.

3
 • Machines can store memory, adapt responses based on prior events, and even “fake”
emotional reactions.

No, machines cannot have emotions

• Emotions require subjective experience—an inner “feeling”—which machines lack.

• Human emotions often involve biochemical processes that machines cannot replicate
exactly.

The instructor played devil’s advocate, repeatedly asking “Why not?” to test the strength of the
“no” position. He pointed out that we ourselves do not fully understand how emotions work in
humans. Without that knowledge, how can we confidently say they cannot exist in machines?

One student suggested that humans experience “multitudes” of emotions simultaneously, which
machines might struggle to do. The instructor countered that nothing in principle prevents a
machine from processing multiple emotional states in parallel, given sufficient design.

The discussion ended with an important philosophical takeaway: our inability to clearly define
and understand emotions limits our ability to assess their presence (or absence) in
machines. The debate remains open.

Defining Consciousness
The conversation shifted to consciousness, often casually equated with “awareness.” The
instructor pushed students to unpack this. What exactly does awareness mean? Is it “knowing
what you are, where you are, and what you’ll do next”? If so, could a machine meet those
criteria?

One student suggested that humans are conscious “all the time.” The instructor challenged this
with an observation exercise: much of our daily activity is automatic—walking, eating, or typing
without deliberate awareness of each movement. Consciousness, in this sense, spikes when we
pause to reflect or direct attention inward.

This tied into the higher-order thought theory of consciousness, which suggests that
consciousness arises when we become aware of our own thoughts. The instructor used relatable
examples: Are you aware right now of the feeling of your shoes on your feet? Probably not—
until it’s pointed out.

By making students aware of their own selective awareness, the instructor set the stage for later
arguments that machines, too, could model self-awareness, at least in functional terms, if
they track and report their own states.

Filtering and Selective Awareness

Drawing on neuroscience, the instructor explained that the human brain processes massive
amounts of sensory input every second but filters out the vast majority. This selective awareness
ensures we focus only on information critical for survival.

4
For example, your clothes are constantly in contact with your skin, but you rarely notice unless
you consciously attend to the sensation. The brain deems such information non-essential and
filters it out to avoid cognitive overload.

The philosophical hook was that this is not uniquely human. Machines already perform similar
filtering—such as AI systems prioritizing relevant data and ignoring noise. In some domains,
they do it more efficiently than we do.

The implication is profound: if selective attention is part of what we associate with

consciousness, then a machine that filters and prioritizes information could arguably possess a
form of it. This erodes one more supposed boundary between human and machine cognition.

By linking neuroscience to AI, the instructor demonstrated that philosophy of mind must
engage with empirical science, since understanding how we filter experience informs debates
about artificial awareness.

Programming – Humans and Machines

The instructor moved to the idea that both humans and machines are “programmed.” In
machines, programming is literal—lines of code written to specify tasks. In humans,
programming is more metaphorical but no less real. From the moment we are born, we are
shaped by our environment, culture, language, and experiences. Parents teach rules (“don’t touch
the stove,” “say thank you”), schools instill habits, and repeated exposure to ideas reinforces
certain patterns of thought.

The instructor pointed out that much of this “human programming” happens unconsciously.
Our brains filter information, strengthen certain neural pathways through repetition, and weaken
unused connections—a process not unlike how machine learning algorithms adjust weights in a
neural network. Conscious self-programming also happens: deciding to learn a new skill,
adopting a moral principle, or changing a habit is deliberate reconfiguration of our own “code.”

The analogy is provocative. If humans are programmed, and machines are programmed, the
difference might be in the type of programming (biological, social, or digital) rather than the fact
of being programmed. This raises philosophical questions about free will: if our responses are
shaped by inputs and conditioning, in what sense are we fundamentally different from
sophisticated machines?

Reproduction – Biological and Artificial

When the class discussed reproduction, a student asserted that machines cannot reproduce. The
instructor immediately challenged this assumption by asking what reproduction actually means.
If it’s defined as the creation of new entities of the same kind, then self-replication is the key—
not biological mating.

A humorous exchange followed: a student said, “Two cars can’t make a car,” implying that
machines lack the biological capacity for reproduction. The instructor countered that
reproduction does not have to involve biological processes. Just as humans can use assembly
lines to build machines, it’s logically possible to design machines capable of building other
machines autonomously.

5
He introduced the idea of logical possibility versus current reality. Today, most machines
cannot self-assemble. But in principle, it’s not impossible. The obstacle is technological, not
conceptual. Historical parallels—such as human flight—show that many things once deemed
impossible became reality with innovation.

This shift in perspective reframed reproduction as a matter of design and resources, not species-
specific biology.

Logical vs. Actual Possibility

This distinction became its own topic. The instructor explained that philosophy often deals with
what is logically possible—scenarios that could happen without contradiction in the laws of
logic—even if they are not actually possible right now due to technological or practical
constraints.

He gave the example of flight: thousands of years ago, it was logically possible for humans to
travel through the air, but it was not practically achievable until the development of aeronautics.
Similarly, autonomous self-replicating machines are logically possible, even if current engineering
cannot produce them reliably.

The value of distinguishing logical from actual possibility lies in clarifying debates. When
someone says “That can’t be done,” do they mean it’s logically impossible (like a square circle) or
just currently impractical? In philosophical and technological discussions, these are very different
claims.

This framing encouraged students to shift from thinking in terms of present limits to thinking
about conceptual limits. In doing so, they could explore the boundaries of possibility without
being trapped by the state of current technology.

Pain, Fatigue, and Human Limits

Finally, the class considered physical and experiential limits. Students suggested that humans can
feel pain and fatigue, while machines cannot. The instructor asked for a definition of pain, noting
that some humans—due to medical conditions like congenital insensitivity—cannot feel pain at
all. They are still human, so the ability to feel pain is not essential to being human.

Similarly, machines can experience operational limits—overheating, system errors, or degraded

performance—which could be analogized to fatigue. While these machine “failures” lack
subjective sensation, the functional role is similar: performance drops when conditions exceed
capacity.

This discussion also touched on error tolerance. Students initially thought humans are “allowed”
to make mistakes while machines are not. But machines do make errors—software bugs,
miscalculations, sensor misreads—often with real-world consequences.

The philosophical takeaway was that many traits used to distinguish humans from machines are
not as clear-cut as they seem. Functional analogues exist, even if the underlying mechanisms
differ. The real challenge lies in deciding whether the difference in how an experience is produced
matters more than the effect it has.

6
Humans as Biological Machines – A Philosophical
Reflection
The session concluded with a provocative but deeply instructive claim: humans are biological
machines. This statement, while seemingly reductive, is grounded in both philosophical
reasoning and scientific understanding. The instructor’s point was not to diminish human
experience, but to caution against unexamined assumptions of human exceptionalism. When we
think humans are special without understanding the fundamentals of what we are, we often
create false dichotomies—divisions between “us” and “machines” that rest more on bias than on
logic.

From the discussions earlier in class, we saw these dichotomies arise repeatedly. Students
proposed that humans are emotional while machines are not (Point 3), creative while machines
are merely programmed (Point 3), and capable of consciousness in a way machines cannot match
(Point 9). But at each step, the instructor encouraged a deeper inquiry into definitions. If
thinking is defined as symbol manipulation (Point 4), then both humans and machines qualify.
If emotions are internal states producing behavioral outcomes (Point 7), then machines
might simulate them functionally, even if their mechanisms differ.

Philosophically, this approach echoes functionalism in the philosophy of mind—a position

defended by scholars such as Hilary Putnam and Jerry Fodor—which holds that mental states
are defined by what they do, not by what they are made of. By this standard, whether the
“machine” is built of neurons and synapses or of silicon and circuits may be irrelevant to the
classification of its processes.

The idea of humans as biological machines is also supported by neuroscience. The brain takes
inputs from sensory organs, processes them through complex neural networks, and produces
outputs in the form of thoughts, speech, and action—mirroring the input-process-output
structure of machines (Point 6). Our learning is a form of programming: repeated exposure
shapes neural pathways much as repeated training adjusts the parameters of an artificial neural
network (Point 11).

One source of bias in thinking about machines is our anthropocentric perspective—the

tendency to use ourselves as the gold standard for intelligence, emotion, and consciousness. In
the session, this bias appeared in the surprise some students expressed when the instructor
suggested bacteria “think” in rudimentary symbolic terms (Point 5). Yet when the concept of
thinking was stripped to its functional essence, the barrier between “high-level” human cognition
and “low-level” bacterial or mechanical processing became more porous.

This human–machine dichotomy is further weakened when we consider logical versus actual
possibility (Point 13). Many capacities we take as uniquely human—communication, tool use,
complex decision-making—are, in principle, replicable in machines. That they are not yet fully
realized mechanically is a technological constraint, not a logical impossibility. History is full of
once-impossible feats becoming mundane, from flight to instant global communication.

Finally, the closing point has an ethical and epistemic dimension: clinging to human
exceptionalism without examining our reasoning risks misunderstanding both ourselves and
the technologies we create. By seeing humans as biological machines, we open the door to a
more accurate and less biased exploration of cognition, emotion, and consciousness—one that
evaluates capacities by function and structure, not by origin.

7
The instructor’s message was not to erase the richness of human life, but to challenge the
intellectual complacency that comes from assuming special status without justification. In
philosophy, this shift from intuition-based dichotomies to concept-based analysis is
essential for honest inquiry. Recognizing ourselves as biological machines allows us to compare,
contrast, and design artificial systems without the distortions of unexamined bias—and, perhaps,
to better understand the remarkable machine that each of us already is.