1.What this chapter is about
This topic pack connects four areas:
artificial intelligence
advertising
memory
psychology
At C1, the goal is not merely isolated vocabulary, but the relations among these areas:
How is attention guided?
How do algorithms personalise content?
How do memory and recognition arise?
How can advertising influence perception and decisions?
What roles do data, expectations, and cognitive biases play?
The aim is precise, nuanced discussion without overstatement.
2.Core concepts
Core vocabulary:
die künstliche Intelligenz (KI)
das Modell
der Algorithmus
die Personalisierung
die Zielgruppe
die Werbung
die Aufmerksamkeit
die Wahrnehmung
das Gedächtnis
die Erinnerung
die Wiedererkennung
die Entscheidung
die Beeinflussung
die Verzerrung
die Erwartung
These concepts form the framework for later discussion.
3.AI is not one single system
The term AI is very broad.
It can refer to different systems:
language models
image generators
recommender systems
classification models
forecasting models
advertising optimisation
automated decision support
A precise C1 discussion should therefore specify what kind of system is meant.
4.Algorithm vs AI
Not every algorithm is AI.
algorithm
→ a general computational or decision rule
AI system
→ often a data-driven model that recognises, generates, or predicts patterns
Example:
A fixed sorting rule is an algorithm,
but not automatically an AI model.
5.Model, data, and training
Important terms:
training data
training process
model parameters
input
output
prediction
error rate
generalisation
Useful phrases:
Das Modell wurde mit ... trainiert.
Die Vorhersage basiert auf ...
Die Ausgabe hängt unter anderem von ... ab.
6.Personalisation
Personalisation means adapting content, recommendations, or ads to characteristics associated with a user.
Typical factors include:
past clicks
search behaviour
purchase history
location category
device type
interest profiles
interaction patterns
Important:
Personalisation is not the same as perfectly knowing a person.
It is based on data and probabilities.
7.Recommender systems
Recommender systems help determine which content becomes visible.
Typical verbs:
Inhalte empfehlen
Beiträge priorisieren
Produkte vorschlagen
Interessen ableiten
Nutzungsverhalten auswerten
Example:
Das System priorisiert Inhalte, die mit hoher Wahrscheinlichkeit relevant erscheinen.
8.Ranking rather than neutral order
Digital content is often not shown in a neutral or random order.
Ranking can determine:
which post appears first
which product is highlighted
which ad is delivered
which video is recommended
Visibility itself therefore becomes an important factor.
9.Generative AI
Generative AI creates new content, for example:
text
images
audio
video
synthetic voices
advertising variants
In advertising, it can be used to:
create variants faster
adapt copy
develop image concepts
test audience messaging
10.Automation in advertising
Advertising processes can be partially automated:
audience segmentation
bid optimisation
ad selection
A/B variants
budget allocation
performance analysis
Useful phrases:
Der Prozess wird teilweise automatisiert.
Das System optimiert auf ...
Die Kampagne wird anhand von ... angepasst.
11.Target audience
A target audience is a group to which communication is directed.
Criteria may include:
age
interests
needs
purchasing behaviour
life stage
occupation
media use
At C1, avoid treating target groups as perfectly homogeneous.
12.Segmentation
Segmentation means:
a broad audience
→ divided into smaller groups
For example:
by usage frequency
by product interest
by purchase stage
by price sensitivity
Potential benefit:
more precise communication
Potential risk:
oversimplification or stereotyped profiles
13.Targeting
Targeting determines who is shown a particular advertising message.
Typical expressions:
eine Zielgruppe ansprechen
eine Anzeige ausspielen
Nutzer gezielt erreichen
Kriterien festlegen
Reichweite begrenzen
Targeting is therefore about audience selection.
14.Personalised advertising
Personalised advertising adapts content to inferred interests or characteristics.
Example:
Two users see different ads
although they use the same platform.
Useful phrases:
Die Anzeige wird an ... angepasst.
Das System berücksichtigt ...
Die Botschaft variiert je nach ...
15.Contextual advertising
Not all targeted advertising relies on personal profiles.
Contextual advertising may be based on:
page content
video topic
search query
current context
Example:
An article about hiking shows ads for outdoor equipment.
16.Retargeting
Retargeting refers to advertising that reaches users again after an earlier interaction.
Typical situations:
viewed a product
abandoned a cart
visited a website
started registration
Example:
After visiting a shop, the user later sees ads for similar products.
17.Advertising message
An advertising message may combine different elements:
product information
value proposition
emotion
social signals
price
scarcity
call to action
C1 phrasing:
Die Botschaft verbindet sachliche Information mit emotionaler Ansprache.
18.Attention as a scarce resource
Digital communication competes for limited attention.
Typical expressions:
Aufmerksamkeit gewinnen
Aufmerksamkeit binden
Aufmerksamkeit lenken
Aufmerksamkeit verlieren
Reize hervorheben
Ablenkung erzeugen
Example:
Short visual cues are intended to direct attention to a central element early.
19.Selective attention
People do not process all available stimuli equally.
Selective attention means that some information receives preferential processing.
Example:
Strong colours or movement can highlight certain elements.
Important:
Visibility can increase attention, but does not guarantee persuasion.
20.Perception
Perception is more than simply seeing or hearing.
It includes:
selection
interpretation
categorisation
attribution of meaning
Useful phrases:
etwas wahrnehmen als ...
einen Eindruck vermitteln
eine Wirkung hervorrufen
unterschiedlich interpretiert werden
21.Priming
Priming refers to effects in which earlier stimuli can influence the processing of later information.
Example:
Certain words or images can make later associations more likely.
Careful phrasing:
Der vorherige Reiz kann die spätere Wahrnehmung beeinflussen.
Not:
The stimulus programs behaviour.
22.Framing
Framing describes how the presentation of the same information can influence interpretation.
Example:
90% fat-free
and:
10% fat
can frame the same fact differently.
Useful phrases:
etwas positiv/negativ rahmen
eine Perspektive hervorheben
einen Aspekt in den Vordergrund stellen
23.Anchoring effect
With anchoring, an initially presented number or piece of information can influence later judgments.
Example:
originally €199
now €129
The first number can act as a reference point.
Careful wording:
Der zuerst genannte Preis kann als Referenzpunkt wirken.
24.Social proof
People often use the behaviour or evaluations of others as a cue.
Advertising signals include:
bestseller
most purchased
more than 10,000 reviews
recommended by many customers
C1 phrasing:
Die Anzeige nutzt soziale Signale, um Vertrauen und wahrgenommene Beliebtheit zu erhöhen.
25.Scarcity
Scarcity can signal urgency.
Typical wording:
only a few left
offer ends today
limited quantity
for a short time only
Critical description:
Die Botschaft erzeugt Zeitdruck, indem sie Knappheit hervorhebt.
26.Loss aversion
Loss aversion describes the tendency to give potential losses substantial weight.
Advertising framing:
Don't miss ...
Secure ...
Don't lose time ...
Careful phrasing:
Die Werbung rahmt die Entscheidung als möglichen Verlust.
27.Emotional appeal
Advertising can appeal to emotions such as:
joy
security
belonging
nostalgia
curiosity
fear
pride
C1 phrases:
Die Kampagne arbeitet mit ...
Die Botschaft spricht das Bedürfnis nach ... an.
Die emotionale Wirkung beruht auf ...
28.Nostalgia and memory
Nostalgic advertising often links products with positive memory-like imagery.
Typical elements include:
music from an earlier period
older designs
childhood scenes
family rituals
familiar brand imagery
This does not necessarily trigger one specific memory; often it creates a memory-like emotional atmosphere.
29.Memory is not a hard drive
Human memory does not operate like a perfect recording device.
It is:
selective
reconstructive
context-dependent
vulnerable to distortion
C1 phrasing:
Erinnerungen werden nicht einfach abgerufen, sondern teilweise rekonstruiert.
30.Short-term and working memory
Working memory can be understood as the limited processing of currently relevant information.
Everyday examples:
holding a phone number briefly
comparing several pieces of information
keeping the beginning of a sentence in mind while listening
For advertising:
Too much information at once can make processing more difficult.
31.Long-term memory
Long-term memory makes information available over extended periods.
Typical domains include:
knowledge
personal experiences
skills
learned associations
Relevant to brand communication:
recognition
familiarity
associations
memorability
32.Encoding
Encoding refers to the processing of new information so that it can enter memory.
Helpful factors may include:
attention
meaning
links to prior knowledge
repetition
emotional relevance
But:
No single factor guarantees lasting memory.
33.Retrieval
Retrieval means making stored information available again.
Typical verbs:
sich an etwas erinnern
etwas abrufen
eine Erinnerung auslösen
etwas wiedererkennen
Example:
A familiar logo can facilitate brand recognition.
34.Recognition vs recall
An important distinction is:
recognition
→ something feels familiar
free recall
→ reproducing information without a direct cue
Example:
recognising a brand logo
≠
recalling the brand name without a cue
Advertising may target either process.
35.Repetition
Repetition can increase familiarity and memorability.
Typical forms:
repeated logo
recurring slogan
same sonic logo
consistent colour palette
regular brand presence
But:
Excessive repetition may also be perceived as intrusive or annoying.
36.Mere-exposure effect
The mere-exposure effect refers broadly to the tendency for repeatedly encountered stimuli to become more familiar.
In advertising:
repeated exposure
→ greater familiarity
→ possibly more positive evaluation
Important:
The effect does not guarantee liking or purchase.
37.Forgetting
Forgetting is not a single process.
It can relate to:
weak encoding
retrieval failure
interference
time delay
low relevance
Useful phrases:
Information verblasst.
Eine Erinnerung wird schwerer zugänglich.
Neue Informationen überlagern ältere.
38.Interference
Similar information can interfere with one another.
Example:
several very similar advertising messages
→ lower distinctiveness
C1 phrasing:
Die Botschaften konkurrieren im Gedächtnis miteinander und können sich gegenseitig überlagern.
39.Context-dependent memory
Memory can be influenced by the context in which information was encoded and later retrieved.
Relevant to advertising:
place
situation
mood
medium
surrounding cues
Careful phrasing:
Bestimmte Kontextreize können den Abruf erleichtern.
40.Emotion and memory
Emotionally relevant events can be especially salient and memorable.
But:
strong emotion
≠
perfect memory
Emotional memories can still be incomplete or distorted.
C1 sentence:
Emotion kann die Erinnerbarkeit erhöhen, ohne die Genauigkeit automatisch zu garantieren.
41.Memory and brand associations
Brands can become linked with particular qualities or situations.
Examples:
security
innovation
luxury
everyday life
family
sport
sustainability
Useful phrases:
Die Marke wird mit ... assoziiert.
Die Kampagne versucht, ... mit der Marke zu verknüpfen.
42.Brand recognition
Typical factors include:
logo
colours
typography
shape
slogan
sound
packaging
recurring visual language
C1 phrasing:
Konsistente Gestaltung kann die Wiedererkennung erleichtern und die Marke visuell stabilisieren.
43.Storytelling in advertising
A story can place information into a structure:
initial situation
problem
conflict
solution
outcome
This can make abstract qualities more concrete.
Example:
Instead of merely saying “reliable”, an ad shows a situation in which reliability matters.
44.Narrative identification
People may identify with characters or situations in a narrative.
Relevant to advertising:
similar life situation
similar problems
similar goals
Careful phrasing:
Die Geschichte erleichtert es bestimmten Zuschauern möglicherweise, sich in der Situation wiederzufinden.
45.Familiarity
Familiarity can influence decisions because familiar information may be easier to process.
Typical words:
vertraut
bekannt
wiedererkennbar
geläufig
But:
Familiarity does not automatically mean quality or truth.
46.Processing fluency
Information that is easy to process may feel more familiar or plausible.
Examples:
clear typography
simple wording
familiar structure
easy-to-see contrasts
Caution:
easier to process
≠
more factually correct
47.Confirmation bias
Confirmation bias refers to the tendency to attend to or weight information more strongly when it fits existing expectations.
This can matter online when:
users prefer similar content
algorithms recommend similar content
existing views are repeatedly reinforced
Careful phrasing:
Dies kann bestehende Überzeugungen stabilisieren.
48.Availability heuristic
People may estimate likelihood partly by how easily examples come to mind.
Advertising or media exposure can therefore increase the mental availability of certain topics.
Example:
If a risk is highly visible, it may feel more present.
That is not the same as an objective increase in risk.
49.Halo effect
With the halo effect, one salient positive or negative feature can influence the overall evaluation.
Advertising examples:
attractive design
famous person
premium visual language
prestigious context
C1 sentence:
Ein einzelnes positives Merkmal kann auf die Gesamtwahrnehmung ausstrahlen.
50.Authority as a cue
Advertising sometimes uses experts, institutions, or public figures as trust cues.
Typical forms:
recommended by experts
scientifically tested
certified
featuring a specialist
Critical question:
Is the authority relevant to the topic, and is the claim verifiable?
51.Influencer advertising
Influencer advertising often combines:
a personal relationship with the audience
everyday context
product placement
recommendation
entertainment
Advertising and personal communication can therefore blend together.
C1 phrasing:
Die Grenze zwischen persönlicher Empfehlung und kommerzieller Kommunikation kann weniger deutlich wirken.
52.Recognising advertising
Media literacy also means being able to recognise commercial communication.
Questions include:
Who finances the content?
What is being promoted?
Is the advertising labelled?
What interests are behind it?
What data may have been used to deliver it?
53.Native advertising
Native advertising is designed to resemble editorial or platform-native content.
Examples:
sponsored article
paid post
integrated product feature
sponsored recommendation
Critical point:
The commercial intention may be less obvious.
54.Dark patterns
Dark patterns are design patterns that can push users toward certain choices or make refusal harder.
Examples:
hard-to-find rejection option
preselected consent
unnecessarily difficult cancellation
misleading buttons
artificial time pressure
C1 phrasing:
Das Interface lenkt die Entscheidung durch asymmetrische Gestaltung.
55.Nudging
Nudging refers to design that influences decisions without fully removing options.
Examples:
default setting
order of options
highlighted recommendation
reminder
Discussion point:
When does a nudge support a decision,
and when does it become manipulative?
56.Use manipulation carefully
Manipulation is a strongly evaluative term.
Before using it, describe:
what information is missing
what pressure is created
which alternative is made harder
what perception is intentionally encouraged
Instead of:
This is manipulation.
more precisely:
Die Gestaltung erschwert eine informierte, symmetrische Wahl.
57.Persuasion vs manipulation
One possible analytical distinction:
persuasion
→ arguments or appeals where meaningful choice remains visible
problematic influence
→ relevant information is hidden or decision freedom is deliberately constrained
The boundary can be contested in individual cases.
58.Data profile
A data profile may combine different types of information.
For example:
interactions
purchases
searches
click patterns
device characteristics
inferred interests
Important:
A profile may contain not only directly provided information, but also inferred characteristics.
59.Explicit vs inferred data
Difference:
explicit
→ the user provides the information directly
inferred
→ the system estimates a characteristic from behaviour
Example:
explicit:
“I am interested in photography.”
inferred:
frequent interaction with photography content
→ inferred interest
60.Profiling
Profiling refers to systematic evaluation or categorisation based on data.
Typical verbs:
categorise users
infer interests
analyse behaviour
estimate probabilities
update profiles
Critical question:
Which inferences are justified, and how transparent are they?
61.Data protection and advertising
Advertising raises questions such as:
What data are collected?
What are they used for?
How long are they stored?
Who gets access?
Is the use understandable?
What choices do users have?
In a language-learning context, careful general phrasing is often preferable to specific legal claims.
62.Consent
Important vocabulary:
die Einwilligung
zustimmen
ablehnen
widerrufen
informiert sein
freiwillig entscheiden
Discussion phrase:
Entscheidend ist nicht nur, ob eine Zustimmung formal eingeholt wird, sondern auch, wie verständlich und frei die Entscheidung tatsächlich ist.
63.Transparency
Transparency can concern several layers:
Why am I seeing this ad?
What data were considered?
Is the content paid?
Is an image synthetically generated?
Who is responsible?
Useful verbs:
transparent machen
offenlegen
kennzeichnen
nachvollziehbar erklären
64.Bias in data and models
Bias can have several meanings.
In AI, it may refer to systematic distortions caused by:
imbalanced training data
poor target variables
historical patterns
incomplete features
measurement error
C1 phrasing:
Das Ergebnis kann bestehende Verzerrungen reproduzieren oder verstärken.
65.Advertising bias vs model bias
It is useful to distinguish:
cognitive bias
→ a human thinking tendency
model bias
→ systematic distortion in a model or dataset
advertising framing
→ deliberate presentation of a message
These concepts are related but not synonymous.
66.Filter bubbles: use the term carefully
The term filter bubble is used when personalised selection leads users to see mostly similar content.
Careful phrasing:
Personalised systems can influence the diversity of visible content.
Too strong:
Algorithms automatically trap people in a filter bubble.
Actual effects depend on the system and user behaviour.
67.Echo chamber
An echo chamber describes a communication environment in which similar views reinforce one another.
Possible factors include:
social selection
group norms
platform logic
recommendations
personal preferences
The cause should not automatically be attributed only to technology.
68.Attention economy
The term attention economy describes situations in which attention is scarce and economically valuable.
Typical verbs:
monetise attention
increase time spent
encourage interaction
maximise reach
C1 sentence:
Plattformen konkurrieren nicht nur um Nutzer, sondern auch um deren Zeit und Aufmerksamkeit.
69.Engagement
In platform contexts, engagement includes interactions such as:
clicks
likes
comments
shares
time spent
But:
High engagement does not automatically mean high quality or satisfaction.
Useful phrases:
auf Engagement optimieren
Interaktion messen
Verweildauer erhöhen
70.Conversion
A conversion is a desired action.
For example:
purchase
registration
download
contact request
booking
Typical expressions:
die Conversion-Rate erhöhen
eine Handlung auslösen
vom Klick zum Kauf
71.A/B testing
A/B tests compare variants.
Example:
variant A:
blue button
variant B:
green button
Possible metrics:
click rate
sign-up
purchase
Important:
A difference in a metric does not automatically reveal the psychological mechanism behind it.
72.Correlation and causation
Typical mistake:
Users saw more ads and bought more.
→ The ads caused the purchase.
That does not automatically follow.
C1 phrases:
Es besteht ein Zusammenhang zwischen ...
Dies deutet auf ... hin.
Eine kausale Wirkung lässt sich daraus allein nicht ableiten.
73.Measuring psychological effects
Possible methods include:
survey
experiment
behavioural data
eye tracking
memory test
reaction time
interview
Each method measures something different.
C1 sentence:
Die Aussagekraft hängt stark davon ab, welche Variable tatsächlich erhoben wurde.
74.Self-report vs behaviour
People can report:
what they remember
what they like
why they bought something
Behavioural data show:
what was actually clicked
viewed
purchased
abandoned
Both data types have different strengths and limitations.
75.Psychology is not mind-reading
Imprecise:
Psychology reveals what people really think.
Better:
Psychological research examines patterns in perception, behaviour, memory, and decision-making under particular conditions.
Psychological claims are also context- and method-dependent.
76.AI is not mind-reading
Imprecise:
The AI knows what I want.
More precise:
The system estimates, based on previous interactions, which content is likely to be relevant to me.
C1 competence means avoiding unnecessary anthropomorphic simplification.
77.Anthropomorphism
Anthropomorphism means attributing human qualities to technical systems.
Typical wording:
The AI wants ...
The AI understands me ...
The algorithm consciously decides ...
More precise:
The model generates ...
The system classifies ...
The algorithm prioritises ...
78.Trust in AI
Trust can be influenced by:
explainability
reliability
experience of errors
brand
user experience
transparency
context
C1 phrasing:
Vertrauen entsteht nicht allein aus technischer Leistung, sondern auch aus der Wahrnehmung von Kontrolle und Nachvollziehbarkeit.
79.Automation bias
Automation bias refers to the tendency to rely too heavily on automated recommendations.
Example:
A recommendation is accepted
despite visible contradictory evidence.
Careful phrasing:
Automated suggestions can reduce the likelihood that people perform independent checks.
80.Algorithm aversion
Conversely, people may reject algorithmic systems quickly after visible errors.
This can create asymmetric evaluation:
human error
→ tolerated
algorithmic error
→ strong loss of trust
Context matters greatly.
81.Explainability
Explainability concerns how well users can understand why a system produced a particular output.
Useful phrases:
eine Entscheidung nachvollziehen
Gründe transparent machen
Einflussfaktoren erläutern
Grenzen offenlegen
Not every model is equally easy to explain.
82.Human in the loop
Human in the loop broadly means retaining human review or decision-making within a process.
Possible functions:
review output
judge borderline cases
correct errors
add context
take responsibility
C1 sentence:
Automatisierung ersetzt nicht zwangsläufig jede menschliche Kontrollinstanz.
83.Synthetic media
AI can generate realistic-looking media:
images
voices
videos
avatars
texts
Discussion questions:
Is the content labelled?
Is the source identifiable?
Can the presentation mislead?
How is trust affected?
84.Deepfakes
A deepfake is synthetic or altered media that can realistically depict real people.
Important verbs:
imitieren
manipulieren
synthetisch erzeugen
kennzeichnen
verifizieren
C1 phrasing:
Realistische Darstellung und tatsächliche Authentizität dürfen nicht gleichgesetzt werden.
85.Advertising with AI-generated people
Advertising may use synthetic people or avatars.
Possible benefits:
flexible production
rapid variation
controlled design
Possible concerns:
transparency
authenticity
expectations about real people
trust
labelling
86.Personalisation vs privacy
A classic trade-off:
more personalisation
→ potentially more relevant content
more data use
→ potentially greater privacy concerns
C1 phrasing:
Der Nutzen personalisierter Angebote muss gegen den Umfang der dafür erforderlichen Datennutzung abgewogen werden.
87.Relevance vs autonomy
Another tension:
the system suggests relevant content
+
the user wants to discover and choose independently
Discussion sentence:
Personalisierung kann Orientierung erleichtern, zugleich aber die sichtbare Auswahl strukturieren und damit indirekt Entscheidungen beeinflussen.
88.Convenience vs control
Automated systems can simplify decisions:
less effort in choosing
faster recommendations
automatic prioritisation
At the same time, users may lose control when:
criteria are unclear
alternatives are hidden
settings are hard to access
89.Responsible psychological language
Psychological topics require careful phrasing.
Better:
can influence
may facilitate
may be associated with
is often explained in terms of
Rather than:
always causes
programs people
controls the subconscious
C1 also means epistemic discipline.
90.Phrase bank: describing a mechanism
Der Effekt beruht darauf, dass ...
Eine mögliche Erklärung ist, dass ...
Dabei spielt ... eine Rolle.
Der Mechanismus lässt sich vereinfacht so beschreiben: ...
Dies kann dazu führen, dass ...
91.Phrase bank: hedging
kann darauf hindeuten, dass ...
dürfte eine Rolle spielen
lässt sich nicht eindeutig feststellen
hängt stark vom Kontext ab
gilt nicht gleichermaßen für alle
sollte nicht überinterpretiert werden
92.Phrase bank: weighing advantages and risks
Ein Vorteil besteht darin, dass ...
Dem steht gegenüber, dass ...
Problematisch wird es insbesondere dann, wenn ...
Aus Nutzersicht spricht für ..., dass ...
Kritisch zu sehen ist ...
Der Nutzen hängt davon ab, ob ...
93.Phrase bank: analysing advertising
Die Anzeige richtet sich an ...
Die Botschaft stellt ... in den Vordergrund.
Die Kampagne arbeitet mit ...
Die Gestaltung lenkt die Aufmerksamkeit auf ...
Die Wirkung entsteht vor allem durch ...
Die Werbung rahmt das Produkt als ...
94.Phrase bank: analysing AI
Das System basiert auf ...
Das Modell wurde mit ... trainiert.
Die Ausgabe hängt unter anderem von ... ab.
Der Algorithmus priorisiert ...
Das System schätzt die Wahrscheinlichkeit, dass ...
Die Entscheidung wird teilweise automatisiert.
95.Phrase bank: talking about memory
Die Information wird leichter erinnert, wenn ...
Die Wiedererkennung wird durch ... erleichtert.
Die Erinnerung ist nicht vollständig zuverlässig.
Neue Informationen können ältere überlagern.
Der Kontext kann den Abruf beeinflussen.
96.Phrase bank: critical evaluation
Dabei stellt sich die Frage, ob ...
Entscheidend ist, inwieweit ...
Kritisch wird es, sobald ...
Nicht eindeutig ist, ob ...
Eine pauschale Bewertung greift zu kurz.
Es kommt wesentlich darauf an, wie ...
97.Common mistake 1: AI knows everything
Too broad:
The AI knows the user exactly.
Better:
The system infers probabilities about interests from observed behaviour.
The second version separates data, inference, and uncertainty.
98.Common mistake 2: advertising always manipulates
Too strong:
Advertising manipulates people.
More precise:
Advertising attempts to influence attention, perception, and decisions; the degree of influence depends on content, context, and the individual.
99.Common mistake 3: emotion equals irrationality
Imprecise:
Emotional decisions are irrational.
Better:
Emotions can be a relevant part of evaluation and decision-making; problems arise when the underlying mechanisms are oversimplified.
100.Common mistake 4: memory as video recording
Imprecise:
Memory stores everything and later retrieves it.
Better:
Memory is selective and reconstructive; memories can change and depend on context.
101.Common mistake 5: bias means simple human error
Bias does not simply mean:
someone thinks incorrectly
It can refer to a systematic tendency.
Also distinguish:
cognitive bias
data bias
model bias
102.Common mistake 6: personalisation equals true individuality
A personalised system may still rely on group-level patterns.
Therefore:
personalised
≠
fully understood as an individual
More precise:
The offer is adapted based on selected characteristics or behavioural data.
103.Common mistake 7: correlation equals mechanism
If a campaign coincides with higher sales, this does not automatically reveal:
which element worked
why it worked
for whom it worked
C1 sentence:
The observed effect does not yet allow a clear conclusion about the underlying psychological mechanism.
104.An algorithm for analysing an advertisement
Practical scheme:
1. identify the target audience
2. state the core message
3. describe visual and verbal techniques
4. cautiously formulate the possible psychological mechanism
5. examine data or personalisation aspects
6. weigh possible benefits and risks
7. do not claim effects that are not supported
105.An algorithm for analysing an AI system
1. What task does the system perform?
2. What data does it use?
3. What output does it produce?
4. How is the output used?
5. What errors are possible?
6. How transparent is the process?
7. What human control remains?
8. What psychological or social effects are conceivable?
106.An algorithm for evaluating a psychological claim
For a strong psychological claim, ask:
What exactly is the effect?
Under what conditions?
How was it measured?
How large is it?
Does it apply to everyone?
Is it correlation or causation?
What alternative explanation exists?
This helps avoid popular-psychology oversimplification.
107.Quick map: AI
AI
→ models · data · training
personalisation
→ recommendations · ranking · targeting
generative AI
→ text · image · audio · variants
critical questions
→ bias · transparency · control · privacy
108.Quick map: advertising
advertising
→ target audience · message · channel
mechanisms
→ attention · framing · scarcity · social cues
digital advertising
→ targeting · retargeting · A/B testing · conversion
risks
→ dark patterns · opacity · excessive profiling
109.Quick map: memory
memory
→ encoding · storage · retrieval
memorability
→ attention · meaning · repetition · context
limits
→ forgetting · interference · reconstruction · distortion
advertising
→ recognition · familiarity · brand association
110.Quick map: psychology
perception
→ selection · interpretation
attention
→ selective · limited
decision
→ heuristics · context · emotion
bias
→ systematic tendencies
important:
influence ≠ control
probability ≠ certainty
111.What should become automatic at C1
You should reliably be able to:
name AI systems precisely
distinguish personalisation from “mind-reading”
describe advertising mechanisms neutrally
phrase psychological effects cautiously
explain memory as reconstructive
distinguish types of bias
separate correlation from causation
weigh benefits and risks
include data and transparency questions
justify strong terms such as manipulation
112.The core idea
The four themes are closely connected:
AI
→ selects, generates, or prioritises content
advertising
→ attempts to influence attention and decisions
psychology
→ describes patterns of perception, decision, and evaluation
memory
→ affects what is recognised, remembered, or forgotten
A strong C1 discussion connects these levels without conflating them.
113.Summary
For the topic AI, Advertising, Memory and Psychology, the most important elements are:
precise terminology
careful causal claims
clear separation of data and interpretation
understanding attention and memory
critical analysis of personalisation
nuanced evaluation of advertising
transparency and privacy questions
Remember:
Algorithms, advertising, and psychological mechanisms can influence behaviour, but this does not imply complete control or a simple universal effect. At C1, influence, probability, context, and evidence should be kept linguistically distinct.