Three complete study companions link questions, original Labs and explained answers across the worlds already open: Philosophy, Life & Mind, and Technology. Each activity is reused from the published Nexus Academy course, not invented as a new specialist-reviewed lesson.
Work through the exercises without an account or scores. Print this page to keep the questions and answer key, or save the offline reading pack for a disconnected journey.
The downloadable pack contains all six public introductory guides and these three explained questions. It needs no connection after saving, but external source links require the internet. The three live interactive Labs are not included.
Philosophy & Meaning · STUDY COMPANION 1
1 · Test the structure of a claim
Distinguish a formally valid inference from the factual truth of its premises.
Explore: Read the guide, then open the logic lab and compare modus ponens with affirming the consequent.
Can an argument's structure be valid when one of its premises is false?
Reveal the explained answer
Yes. Formal validity asks whether the conclusion would have to follow if the premises were true. A separate factual inquiry is needed to establish whether the premises are actually true.
Answer key: Yes. Formal validity asks whether the conclusion would have to follow if the premises were true. A separate factual inquiry is needed to establish whether the premises are actually true.
Why the next world connects: A truth table offers a precise formal test. The next world shows how a transparent threshold model can teach a concept without representing every detail of a biological system.
Does the sandbox calculate actual biological voltage or predict an individual neuron's activity?
Reveal the explained answer
No. The sandbox applies a deliberately invented arithmetic rule to invented units. Actual neural signaling involves voltage, ion channels, timing and many interacting biological properties.
Answer key: No. The sandbox applies a deliberately invented arithmetic rule to invented units. Actual neural signaling involves voltage, ion channels, timing and many interacting biological properties.
Why the next world connects: Both the neuron exercise and a software classifier can use thresholds. That similarity does not mean the biological cell and the computing model work in the same way.
Can a high accuracy percentage alone tell you whether the model detects most positive cases?
Reveal the explained answer
No. Accuracy combines all correct predictions. Recall describes how many actual positives were found; the confusion counts also show missed cases and false alarms. Context and the costs of different errors matter.
Answer key: No. Accuracy combines all correct predictions. Recall describes how many actual positives were found; the confusion counts also show missed cases and false alarms. Context and the costs of different errors matter.
Why the next world connects: The three activities illustrate a broader habit: state the precise question, choose a method with transparent assumptions, then name what the result does not prove.
Six distinct before-and-after reading exercises help you examine the meaning, evidence and limits of every introductory guide that is already public. Each links to a second, published question through an existing explained knowledge-graph relationship. These are original study prompts, not independent specialist review, a graded test or a new factual research claim.
Use the expandable worked reflections, print this page, or keep the complete notebook with all six guide chapters in the existing offline reading pack.
Before reading: Write one sentence defining the question of subjective experience. What would count as a conceptual answer, and what could be measured instead?
After reading: Locate a distinction in the published guide between describing brain activity and explaining first-person experience. Which additional question remains after a measurement?
Your notes: _________________________________________
Reveal a possible explanation
The guide treats a description of experience and an empirical account of cognition as related but different tasks. A brain measurement can contribute evidence about associated processes without, by itself, settling every philosophical question about experience.
Worked reflection: The guide treats a description of experience and an empirical account of cognition as related but different tasks. A brain measurement can contribute evidence about associated processes without, by itself, settling every philosophical question about experience.
Next question: When you enter Brain and Cognition, identify what it investigates rather than assuming that a shared word means the two subjects have the same method.
Before reading: List two things a biological explanation of cognition might investigate. Which would require an observation rather than a definition?
After reading: Find where the guide distinguishes studying physical mechanisms from interpreting reports or theoretical models. Write down the difference.
Your notes: _________________________________________
Reveal a possible explanation
This material treats scientific work on biological cognition as evidence about mechanisms and processes. An analogy with a software model may be useful to ask questions, but similarity in vocabulary is not a demonstration of identical physical mechanisms or subjective experience.
Worked reflection: This material treats scientific work on biological cognition as evidence about mechanisms and processes. An analogy with a software model may be useful to ask questions, but similarity in vocabulary is not a demonstration of identical physical mechanisms or subjective experience.
Next question: What evidence would be needed before transferring a conclusion about biological cognition to an artificial system?
Why the published knowledge graph connects these guides
Studies of biological cognition and evaluations of artificial systems use different evidence, so shared vocabulary does not prove identical mechanisms or experience.
RESEARCH NOTEBOOK · 03 · Technology & Futures
Distinguish performance evidence from claims about experience
Before reading: Suppose a program gives convincing answers. What question about its behaviour can be tested, and what philosophical question would remain?
After reading: Find how the published guide distinguishes debates about consciousness from measured AI system performance.
Your notes: _________________________________________
Reveal a possible explanation
A task-performance benchmark supplies evidence about observed outputs under chosen test conditions. It does not, without an additional argument and appropriate evidence, establish that an artificial system has subjective experience.
Worked reflection: A task-performance benchmark supplies evidence about observed outputs under chosen test conditions. It does not, without an additional argument and appropriate evidence, establish that an artificial system has subjective experience.
Next question: When you visit Model Evaluation, which metrics answer a narrower, measurable question than consciousness?
Before reading: Write down an 'if P then Q' statement. Can the conclusion be true when P is false? Try a counterexample before choosing a rule.
After reading: Use the existing logic Lab to compare modus ponens and affirming the consequent. Look for a true-premises, false-conclusion assignment.
Your notes: _________________________________________
Reveal a possible explanation
A deductively valid argument has no such counterexample under its stipulated form. That logical guarantee is conditional on its premises; it does not independently verify them. The Lab lets readers inspect complete truth-table assignments.
Worked reflection: A deductively valid argument has no such counterexample under its stipulated form. That logical guarantee is conditional on its premises; it does not independently verify them. The Lab lets readers inspect complete truth-table assignments.
Next question: Why is a software classifier's observed error count a different kind of test from a deductive counterexample?
Before reading: What is the difference between a number in an invented model and a measurement of a physical process? Write down your expectations.
After reading: Change the toy Lab's excitation and inhibition values. Compare its arithmetic result with the biological processes discussed in the published guide.
Your notes: _________________________________________
Reveal a possible explanation
The Lab subtracts invented inhibition from invented excitation and compares the result with a chosen boundary. That demonstrates the behaviour of its arithmetic rule, not a prediction about membrane voltage, ion channels, timing or individual neural signals.
Worked reflection: The Lab subtracts invented inhibition from invented excitation and compares the result with a chosen boundary. That demonstrates the behaviour of its arithmetic rule, not a prediction about membrane voltage, ion channels, timing or individual neural signals.
Next question: In the wider Brain and Cognition guide, what additional biological and cognitive processes would an explanatory model need to address?
Before reading: A classifier is reported as accurate. Before accepting that claim, what would you want to know about the test cases and the costs of different mistakes?
After reading: Move the threshold in the fictional classifier Lab and record false positives and false negatives. Compare accuracy with recall.
Your notes: _________________________________________
Reveal a possible explanation
Accuracy combines correct classifications, while recall describes how many actual positive examples were detected. On the Lab's fictional set, changing the threshold can change the trade-off. None of these numbers alone guarantees real-world safety or generalisation.
Worked reflection: Accuracy combines correct classifications, while recall describes how many actual positive examples were detected. On the Lab's fictional set, changing the threshold can change the trade-off. None of these numbers alone guarantees real-world safety or generalisation.
Next question: How does this experimental comparison differ from a proof that a conclusion follows from premises?
Why the published knowledge graph connects these guides
Both use explicit rules to test whether conclusions follow from information, though formal validity and predictive performance are different concepts.
M27 · EVIDENCE DETECTIVE · SIX ORIGINAL SCENARIOS
Six decisions: what does the evidence really show?
Choose one interpretation in each fictional scenario, then open the choice to read an immediate explanation. Every case draws on a different published guide and connects to another through an already-published knowledge-graph relationship. No scores, accounts, data collection, scripts or external services are needed.
These situations are original teaching examples—not real research, diagnostic advice, verified field data or claims of independent specialist approval. They teach the difference between observation, inference and the limits of a model.
CASE 01 · Formal reasoning · Philosophy & Meaning
A conclusion that sounds convincing
Fictional situation: A fictional researcher proposes: If a sample is from a particular archive, its label is blue. This sample has a blue label. The researcher concludes that it must come from that archive. Several other archives also use blue labels, but the researcher has not investigated them.
Your evidence decision: Which conclusion follows from the premises as a matter of deductive logic rather than from a reassuring resemblance?
Option A: The source archive follows necessarily from the blue label.
Not supported by the given evidence: The argument affirms the consequent. If the archive implies a blue label, a blue label does not establish the archive; another archive could have the same label.
Option B: The inference is invalid; a blue label alone leaves other possibilities open.
Supported interpretation: Correct. An assignment in which the label is blue but the sample is from another archive makes the premises true and the conclusion false.
Option C: The claim is invalid only if the blue label was measured incorrectly.
Not supported by the given evidence: Measurement accuracy is a different question. Even a correctly observed blue label cannot turn this logical form into a valid deduction.
Read the full methodological explanation
A deductively valid argument forbids true premises and a false conclusion in its stated form. Here a counterexample is possible, so the form fails independently of whether a real archive is likely. Contrast this with measuring how often a classification rule succeeds on chosen cases.
What this example cannot establish: This is an invented archive example. No real provenance research, statistics or historical evidence has been supplied, and the exercise should not be used to assess any actual collection.
Worked answer: Option B. A deductively valid argument forbids true premises and a false conclusion in its stated form. Here a counterexample is possible, so the form fails independently of whether a real archive is likely. Contrast this with measuring how often a classification rule succeeds on chosen cases.
Continue the question: When you move to Model Evaluation, ask how a false positive in a finite test set differs from a counterexample to logical entailment.
Why these guides connect: Both use explicit rules to test whether conclusions follow from information, though formal validity and predictive performance are different concepts.
CASE 02 · Philosophical interpretation · Philosophy & Meaning
What did the observation actually settle?
Fictional situation: In a fictional study, participants describe seeing a moving shape while researchers record a change in one measured signal. The descriptions and recordings tend to occur together. An enthusiastic report says the recording has fully solved what subjective experience is.
Your evidence decision: Which interpretation best respects the boundary between a correlation about cognition and a broader philosophical question?
Option A: The associated signal alone settles every question about subjective experience.
Not supported by the given evidence: A recorded association can be informative without answering all philosophical questions about what an experience is or why it feels a particular way.
Option B: The observation is useless because experience is always subjective.
Not supported by the given evidence: Empirical observations can contribute important constraints to theories. The error is not studying the brain; it is claiming more than the evidence supports.
Option C: The association may inform a theory, but further explanations and definitions are needed.
Supported interpretation: Correct. The findings might constrain accounts of cognition while leaving open questions about explanation, interpretation and whether an association is causal.
Read the full methodological explanation
The philosophy of consciousness can ask which concept of experience is being studied, while neuroscience can investigate measurable processes linked to reports or behaviours. Neither method is interchangeable with the other simply because both use the word consciousness.
What this example cannot establish: No human participants were studied for this exercise, and no neuroscience finding or clinical conclusion is being reported. The imagined observation does not establish how actual brain activity causes experience.
Worked answer: Option C. The philosophy of consciousness can ask which concept of experience is being studied, while neuroscience can investigate measurable processes linked to reports or behaviours. Neither method is interchangeable with the other simply because both use the word consciousness.
Continue the question: In Brain and Cognition, identify what a real experimental design would measure before translating an observation into a conceptual explanation.
Why these guides connect: Scientific study of cognition offers a different perspective on questions of subjective experience.
CASE 03 · Evidence about learning · Life, Body & Mind
The familiarity trap
Fictional situation: A fictional learner rereads the same passage repeatedly and begins to recognise its phrases quickly. The next day, the learner struggles to explain a new example without the passage. A mentor declares that the fast recognition proved complete understanding.
Your evidence decision: What would be a more informative way to test the narrow claim that the learner can explain and apply the idea?
Option A: Use a delayed explanation task with a different example and transparent criteria.
Supported interpretation: Correct. A different application task can test more than familiarity, although the task still needs fair conditions, appropriate feedback and attention to prior knowledge.
Option B: Count the number of times the same passage is reread and call that a comprehension score.
Not supported by the given evidence: Repetition count describes an activity, not whether understanding transfers. Repeated exposure can make phrases familiar without establishing application.
Option C: Ask whether the learner feels confident and treat the answer as a biological measurement.
Not supported by the given evidence: Self-reported confidence can be a useful personal reflection, but it is not a direct biological measurement and cannot independently verify learning.
Read the full methodological explanation
Familiarity, retrieval and transfer are different ideas. The stronger assessment depends on the question, timing and task design, rather than treating a single felt sense of fluency as proof of comprehension. The neighbouring neuron guide addresses a different scale of explanation.
What this example cannot establish: This fictional case is not a validated educational intervention, diagnostic test or prescription for any learner. It provides no measured cognitive outcomes, and real accessibility or learning needs may require different approaches.
Worked answer: Option A. Familiarity, retrieval and transfer are different ideas. The stronger assessment depends on the question, timing and task design, rather than treating a single felt sense of fluency as proof of comprehension. The neighbouring neuron guide addresses a different scale of explanation.
Continue the question: When reading Neuron Signals, ask why an illustrative input threshold cannot by itself establish what a whole learner knows.
Why these guides connect: Neural signalling is one of the biological mechanisms studied when investigating cognition and brain activity.
CASE 04 · Model interpretation · Life, Body & Mind
Four invented units do not make a neuron
Fictional situation: A classroom sandbox subtracts an invented inhibition value from an invented excitation value. If the result reaches four units, the interface displays an illustrative signal. With excitation six and inhibition two, the sandbox reports a signal. A visitor says this predicts what a person's brain will do tomorrow.
Your evidence decision: What does the sandbox calculation actually support, given that the inputs and boundary are intentionally invented?
Option A: A clinical prediction about a person's future neural activity.
Not supported by the given evidence: The inputs have no correspondence to an individual's observed voltage, timing, ion channels or biological state. This is not a clinical instrument.
Option B: A demonstration of an arithmetic boundary within a stated fictional model.
Supported interpretation: Correct. Six minus two equals four, so the specified comparison is satisfied. That confirms the model's internal rule—not a real biological event.
Option C: Evidence that every neuron uses the same numerical threshold in these units.
Not supported by the given evidence: The number four is an invented educational boundary. No actual neuron measurements or universal biological claim support that conclusion.
Read the full methodological explanation
The question is not whether the classroom arithmetic is performed correctly but whether the model's definitions match the phenomenon being discussed. A toy rule can help readers inspect a modelling assumption while leaving the biology of real cells unexplained.
What this example cannot establish: These are illustrative units and fictional inputs, not measurements of membrane voltage or patient health. Actual neural signalling involves processes and timescales outside this exercise.
Worked answer: Option B. The question is not whether the classroom arithmetic is performed correctly but whether the model's definitions match the phenomenon being discussed. A toy rule can help readers inspect a modelling assumption while leaving the biology of real cells unexplained.
Continue the question: In the wider Brain and Cognition guide, identify which biological and cognitive mechanisms an elementary threshold equation leaves out.
Why these guides connect: Neural signalling is one of the biological mechanisms studied when investigating cognition and brain activity.
CASE 05 · AI evaluation · Technology & Futures
The appealing accuracy headline
Fictional situation: A fictional classifier correctly labels ninety of one hundred invented observations. Its summary advertises ninety percent accuracy but does not show how many important positive cases it misses. A visitor proposes using that single percentage as proof that the tool is safe in every setting.
Your evidence decision: Which next question would most directly test whether the accuracy headline hides a relevant pattern of classification mistakes?
Option A: Whether the classifier has an attractive interface and sounds confident.
Not supported by the given evidence: Presentation quality may affect usability but says nothing about the distribution or consequences of false negatives and false positives.
Option B: Whether the model is conscious, because consciousness guarantees correct classification.
Not supported by the given evidence: There is no basis for this inference. Subjective experience is not a substitute for evaluating observed errors and deployment conditions.
Option C: What the confusion counts, positive-case recall and test conditions reveal.
Supported interpretation: Correct. The class distribution and error types are essential context; ninety correct predictions can coexist with serious failures on cases that matter.
Read the full methodological explanation
Accuracy aggregates successes, while precision and recall answer more specific questions about positive predictions. Evaluation requires a stated target, documented test population and consequences of mistakes. Even these metrics do not establish how a deployed system will behave under different conditions.
What this example cannot establish: The ninety-of-one-hundred example is fictional, not a benchmark of a real model or an indication of safety or fairness. No deployment or risk-management recommendation follows from this scenario alone.
Worked answer: Option C. Accuracy aggregates successes, while precision and recall answer more specific questions about positive predictions. Evaluation requires a stated target, documented test population and consequences of mistakes. Even these metrics do not establish how a deployed system will behave under different conditions.
Continue the question: When moving to AI and Consciousness, ask why observed task performance does not resolve claims about artificial subjective experience.
Why these guides connect: Measured prediction accuracy does not on its own establish whether a system has subjective experience.
CASE 06 · Concepts and evidence · Technology & Futures
A convincing answer from a machine
Fictional situation: A fictional chatbot writes a moving account of what it might feel like to be lost in a city. Readers praise the writing as persuasive. A commentary concludes that the apparent emotional fluency proves the program has exactly the same kind of subjective experience as a person.
Your evidence decision: What is the most defensible conclusion based only on the output described, without adding unsupported premises about inner experience?
Option A: The system produced persuasive text; claims about subjective experience need further argument.
Supported interpretation: Correct. The visible output can be evaluated as text and behaviour, but it does not by itself settle whether subjective experience exists or how it should be defined.
Option B: A fluent emotional description proves an identical human-like conscious state.
Not supported by the given evidence: This leaps from an observed linguistic output to an unobserved philosophical and empirical claim. The needed inferential bridge has not been established.
Option C: Because experience is disputed, measuring any output has no scientific value.
Not supported by the given evidence: Outputs and task performance can be studied meaningfully. The limitation is treating behavioural success as a complete answer to a different question.
Read the full methodological explanation
AI performance and philosophical consciousness debates overlap in vocabulary but use distinct evidence standards. Evaluating what a system says is one empirical task; interpreting whether a system has experience demands explicit concepts, competing arguments and appropriate additional evidence.
What this example cannot establish: This is a completely fictional program and conversation, not an empirical AI benchmark, a consciousness test or a claim about any existing provider's systems.
Worked answer: Option A. AI performance and philosophical consciousness debates overlap in vocabulary but use distinct evidence standards. Evaluating what a system says is one empirical task; interpreting whether a system has experience demands explicit concepts, competing arguments and appropriate additional evidence.
Continue the question: Revisit Consciousness and ask which different accounts of subjective experience might interpret the same visible behaviour in different ways.
Why these guides connect: Debates about artificial consciousness draw on philosophical definitions and theories of mind.
Your privacy matters
Optional Google Analytics, Tag Manager and Microsoft Clarity scripts load only if you accept. Essential site functions do not require these tags. You can change your choice later.