7.1 Media Literacy and Fake News

Media literacy -- forstå makta til media

Vi lever i ein informasjonsflaum. Nyheiter, meiningar, reklame og propaganda vert blanda saman i sosiale medium, nettaviser og podkastar. Media literacy (mediekompetanse) er evna til å forstå, vurdere og skape medieinnhald kritisk.

Dette kapittelet byggjer vidare på kapittel 1.5 (Digital Literacy) og går djupare inn i:
- Korleis medium formar oppfatninga vår av verda
- Ulike typar feilinformasjon og propaganda
- Algoritmar og filterbobler
- Korleis vere ein kritisk mediebrukar

Filter Bubbles and Echo Chambers

Filter bubble (filterboble):
- Algoritmar på sosiale medium viser deg innhald du sannsynlegvis vil like
- Du får sjå meir av det du allereie er einig i
- Resultatet: eit forvridd bilete av røynda

Echo chamber (ekkokammer):
- Omgir deg med folk som tenkjer likt
- Meiningane dine vert forsterka utan motargument
- Du kan tru at "alle" er einige med deg

Konsekvensar:
- Polarisering -- folk med ulike syn forstår ikkje kvarandre
- Spreiing av feilinformasjon innan lukka grupper
- Vanskelegare å ha nyanserte debattar

Kva kan du gjere?
- Følg ulike nyheitskjelder -- ikkje berre dei du er einig med
- Ver medviten om at algoritmar påverkar kva du ser
- Oppsøk motargument aktivt
- Diskuter med folk som tenkjer annleis enn deg

Social Media Algorithms -- why you see what you see

Every time you open TikTok, Instagram, YouTube, or any social media app, an algorithm decides what you see. But how do these algorithms work, and why should you care?

How algorithms work:
1. Data collection: The platform tracks everything you do -- what you click, how long you watch, what you like, share, and comment on
2. Profiling: Based on this data, the algorithm builds a profile of your interests, preferences, and even your emotional state
3. Content selection: The algorithm shows you content that it predicts you will engage with (click, watch, share)
4. Feedback loop: Your engagement confirms the algorithm's prediction, and it shows you more of the same

Why this matters:
- Engagement over accuracy: Algorithms prioritise content that keeps you scrolling, not content that is true or balanced. Research shows that false stories spread six times faster than true ones on social media, because they trigger stronger emotional reactions (shock, anger, fear)
- Radicalisation pathways: If you watch one extreme video, the algorithm may recommend more extreme content, gradually pushing you towards radical viewpoints
- Emotional manipulation: Algorithms learn what triggers your emotions and exploit this. Content that makes you angry or anxious keeps you scrolling longer
- Addiction by design: Features like infinite scroll, autoplay, and notification badges are designed to keep you on the platform as long as possible

What platforms don't want you to know:
- Internal research at Meta (Facebook/Instagram) showed that Instagram makes body image issues worse for 1 in 3 teenage girls -- but they didn't make the research public until a whistleblower leaked it in 2021
- YouTube's recommendation algorithm has been shown to lead users from mainstream political content to conspiracy theories within a few clicks

Fact-Checking Tools and Deepfakes

How to fact-check

When you encounter a claim online, use these tools and strategies:

Free fact-checking websites:
- Snopes.com -- one of the oldest fact-checking sites, debunks myths and viral claims
- FactCheck.org -- focuses on political claims, run by the University of Pennsylvania
- Full Fact (UK) -- independent fact-checking organisation
- Faktisk.no (Norway) -- Norwegian fact-checking site, useful for local claims

The SIFT method (from chapter 1.5, repeated here because it's so important):
- Stop -- don't immediately react or share
- Investigate the source -- who is behind this information?
- Find better coverage -- do established news outlets report the same thing?
- Trace claims -- find the original source of the claim

Reverse image search:
- Right-click an image and select 'Search image with Google' (or use TinEye.com)
- This reveals if the image has been used before in a different context
- Many fake stories use real photos taken out of context

Deepfakes -- seeing is no longer believing

A deepfake is a video, image, or audio clip created using artificial intelligence that looks or sounds real but is entirely fabricated.

How deepfakes work:
- AI analyses thousands of images/recordings of a person
- It learns to replicate their face, voice, and mannerisms
- It can then create convincing videos of that person saying or doing things they never actually said or did

Real examples:
- Deepfake videos of politicians making false statements have been used to spread disinformation
- Deepfake audio has been used to scam companies out of money by imitating CEO voices
- Celebrities' faces have been placed on other people's bodies without consent

How to spot deepfakes:
- Unnatural blinking or eye movement
- Weird lighting or shadows on the face
- Lip movements slightly out of sync with audio
- Blurry or distorted edges around the face
- The person says something completely out of character

The bigger problem: As deepfake technology improves, it becomes harder to spot fakes. This creates a liar's dividend -- people can dismiss real evidence by claiming it's a deepfake.

Oppsummering

Typar villeiande innhald: fake news, misinformation, disinformation, propaganda, clickbait, deepfake

Filter bubbles og echo chambers: Algoritmar viser deg det du allereie likar, noko som kan gi eit forvridd bilete av røynda.

Ver ein kritisk mediebrukar:
- Sjekk kjelda (SIFT-metoden frå kap. 1.5)
- Følg ulike nyheitskjelder
- Ver medviten om påverknaden frå algoritmar
- Oppsøk motargument
- Ikkje del utan å sjekke først

I ei verd med uendeleg informasjon er evna til å skilje sant frå usant meir verdifull enn nokon gong.