Expand all Collapse all Think Before Sharing Can You Trust What Found You Online? Critical thinking in the age of social media and AI is no longer just the ability to spot a false headline. It is the habit of asking how a claim reached you, who benefits from your reaction, what evidence would change your mind, and whether a machine-generated answer is being treated as a shortcut for judgement. 175 pages Read more Accuracy Nudge Can One Pause Stop a False Share? Small prompts to think about accuracy can help people pause before amplifying a claim that feels urgent or identity-confirming. 7 pages +Show subtopics Read more Evidence How Much Do Accuracy Prompts Actually Help? Across many experiments, accuracy prompts modestly improved the quality of intended news sharing, mainly by reducing false-headline sharing. Read more Nudge Design How to Ask About Accuracy Without Scolding A good nudge is short, neutral, timely, and targeted enough that it adds useful friction without becoming another banner to dismiss. Read more Best Timing The Best Moment to Slow a Viral Claim Accuracy prompts are most useful when they appear just before a user shares, quote-posts, forwards, or otherwise expands a claim's reach. Read more Notes Timing What Community Notes Teach About Acting Early Community Notes can reduce the spread of misleading posts after context appears, but delayed notes show why pre-share nudges still matter. Read more Backfire Risk When More Attention Does Not Mean Better Judgement Designs that make news posts more noticeable can raise engagement without improving truth discernment, especially when attention is not aimed at... Read more Sharing Gap Why We Share Claims We Doubt People often value accuracy but still share false headlines when social rewards, urgency, or group identity take over at the sharing moment. Read more AI Tutors Should You Trust a Chatbot Tutor? Using AI like a search engine or tutor can be useful, but it changes how users should check accuracy and uncertainty. 7 pages +Show subtopics Read more Grounded AI Can Grounded AI Still Be Wrong? Retrieval systems can reduce hallucinations, but they can still misread, overstate or omit important limits in the documents they use. Read more Citation Checks Do AI Citations Prove Anything? A citation beside an AI answer may point to a real page while still failing to support the exact claim being made. Read more Overreliance When AI Tutoring Becomes a Shortcut A chatbot tutor can weaken learning when trust, convenience or quick answers replace struggle, recall and self-correction. Read more Tutor Help When Do Chatbot Tutors Actually Help? AI tutors work best when they guide practice, feedback and explanation instead of simply handing over answers. Read more Source First When Should You Search Instead? Some questions are safer to start with search results because the source trail matters more than a smooth explanation. Read more False Confidence Why Wrong AI Answers Sound Right Chatbots can produce wrong answers in a fluent tone because training can reward plausible guessing over honest uncertainty. Read more AI Virality Why AI Misinformation Travels So Easily AI-made misleading posts can be entertaining, scalable, and surprisingly viral even when they are not the most believable claims online. 7 pages +Show subtopics Read more Pope Puffer The fake puffer jacket that tested everyone The viral Pope puffer image shows how a playful fake can become a public test of visual verification. Read more Spread vs Belief Viral does not mean believed A post can shape attention and memory even when many sharers are not fully convinced it is true. Read more Notes Data What Community Notes can and cannot prove Community Notes data shows what users flagged as misleading, but it cannot capture every AI-made falsehood on X. Read more AI Detectors Why AI detection keeps getting harder As synthetic media improves, detection tools can become less reliable, making reader habits more important. Read more Funny Fakes Why funny AI fakes travel so fast Funny AI-made posts can travel as jokes or visual surprises before readers decide whether they are true. Read more Small Accounts Why small accounts can make big fakes AI tools let small accounts create highly shareable posts that can outrun follower-count expectations. Read more Community Notes Can the Crowd Correct the Feed? Crowd-sourced correction can add context to misleading posts, but readers still need to judge speed, coverage, and source quality. 7 pages +Show subtopics Read more Fact Checks Can crowd notes replace fact checkers? Community Notes and professional fact-checkers often correct different posts, and the strongest public corrections may need both. Read more Bridging Can disagreement make notes more trustworthy? Community Notes depend on agreement across contributors with different rating histories, not just a simple majority vote. Read more AI Scale Can human notes keep up with AI fakes? AI tools can create plausible false images and claims faster than human note systems can reliably review them. Read more Source Check How to read the source behind a note A helpful note is only as strong as the source trail it gives readers for checking the original claim. Read more Late Notes Why corrections often arrive after the damage Community Notes can reduce sharing once visible, but the fastest spread often happens before contributors reach agreement. Read more No Note Why silence under a post proves nothing The absence of a public note may reflect delay, disagreement, low visibility or contributor limits rather than accuracy. Read more Corroboration Who Else Can Confirm This Claim? Urgent claims about disasters, politics, war, health, or crime should be checked against independent sources before sharing. 7 pages +Show subtopics Read more Crisis Rules Can platforms slow rumours in a crisis? Platform crisis rules aim to slow harmful rumours when normal moderation is too slow for fast-moving public emergencies. Read more Safety Claims Checking danger before sharing warnings For crime and emergency posts, useful confirmation usually comes from sources close to the event and accountable for corrections. Read more Southport Case When false identity claims become dangerous False claims after the Southport killings showed how misidentified suspects can move from online rumour to public disorder. Read more Risk Triggers When one source is not enough Some posts need more checking because they ask people to blame, panic, donate, buy, avoid a place, or share immediately. Read more Source Trail Who really started the claim? The strongest check is often not another share, but finding who first made the claim and what evidence they actually had. Read more False Echoes Why many posts can mean one source Many matching posts can still be one weak claim repeated through screenshots, reposts, summaries, and influencer retellings. Read more Deepfakes How to Check a Voice or Video Claim Synthetic voices and videos require new habits for checking provenance, corroboration, and the original recording context. 7 pages +Show subtopics Read more Video Fraud Seeing Them on a Call Is Not Enough A video meeting can feel socially real, but unusual financial requests still need independent approval channels. Read more Credentials What Provenance Labels Really Prove Provenance labels can show parts of a file's history, but they cannot prove every claim attached to the media. Read more Biden Robocall What the Biden Robocall Case Teaches The New Hampshire robocall showed why investigators trace telecom routes and campaign links instead of relying on listeners' ears. Read more Voice Scams When a Loved One's Voice Is Not Proof A familiar voice is not enough proof when a caller creates panic, urgency or pressure to send money. Read more First Upload Where Did That Deepfake Clip Start? The first upload trail often tells more than the clip itself, especially when reposts strip context and reward speed. Read more Detector Scores Why AI Detectors Are Not Verdicts Detector results can help triage suspicious media, but new generators, compression and unfamiliar formats can break confidence. Read more Emotional Posts Why Outrage Is Not Evidence Posts that provoke anger, fear, pride, or moral certainty deserve extra caution because emotional force is not evidence. 7 pages +Show subtopics Read more Social rewards How platforms reward louder moral posts Likes and shares can teach users that more outraged posts get more attention, making emotional escalation feel normal. Read more Pride claims The flattering posts that bypass doubt Pride-based misinformation feels safer than outrage, but it can still make readers accept claims because they flatter a group identity. Read more Moral pressure When checking facts feels socially risky Some posts make asking for evidence seem immoral, which is why scepticism needs to separate truth, framing, and action. Read more Fear urgency When fear turns speed into a trap Fear posts can seem protective in a crisis, but urgent instructions often push weak claims ahead of reliable information. Read more Outrage Why outrage feels like proof online Outrage can turn sharing into a loyalty signal before readers have checked whether the claim is true or fair. Read more AI polish Why polished AI fakes feel more credible AI tools can make emotional misinformation look professional, local, and confident even when the underlying evidence is thin. Read more Evidence Types Not All Evidence Deserves Equal Weight Studies, filings, datasets, regulator notices, named experts, leaks, and anonymous posts carry different kinds of reliability. 7 pages +Show subtopics Read more Viral Charts Before you trust the chart A chart can make a pattern look settled, but the dataset, sampling and methods decide whether the pattern is real. Read more Community Notes Can crowdsourced context be trusted? Community Notes can add context to misleading posts, but their trust, speed, coverage and stability are different questions. Read more Eyewitness Posts The first clip is not the full story Eyewitness clips can reveal real events early, but time, place, angle and reposting can all change what they seem to show. Read more Regulator Notices What a warning letter really proves Regulator warnings are strong evidence of official concern, but they rarely settle every fact or future outcome on their own. Read more Official Records When screenshots are not enough Corporate and legal claims need primary records because screenshots can hide context, amendments and what a filing actually proves. Read more Data Voids When search results go empty When reliable results are scarce, the first confident-looking claim can dominate before better evidence appears. Read more Fake Authority When Official Looking Posts Are Not Official Fake council notices, invented quotes, and official-looking graphics exploit trust in institutions without providing real provenance. 7 pages +Show subtopics Read more Quote Cards Did They Really Say That Locally? A quote card is not proof that a councillor, police source or school official said what the post claims. Read more Safety Warnings Pause Before Sharing Emergency Screenshots Police-like or council-like warnings should be checked through current official channels before they are shared in a crisis. Read more Fake Notices When a Council Notice Is Not Real A council-style planning notice can look decisive, but the real test is whether the application exists in the public record. Read more Parking Scams When a Fine Message Borrows Authority Fake penalty messages exploit the pressure of official enforcement, so payment links and appeal routes need checking before action. Read more Correction Lag Why Local Corrections Arrive Too Late False local notices can travel far before a council, police force or local journalist has time to correct them clearly. Read more Screenshots Why Screenshots Make Rumours Look Official Screenshots strip claims from their source path, making reposted emails, notices and quote cards feel more official than they are. Read more Hallucinations Why Fluent AI Answers Still Need Checking AI systems can produce confident, polished answers that include errors, fabricated context, or sources that do not support the claim. 7 pages +Show subtopics Read more Fluency Check Do Not Mistake Fluency for Evidence A smooth AI answer should be treated as a starting point until its key claims can be traced to reliable outside evidence. Read more Search Summaries How Search AI Turns Snippets Into Certainty Search summaries can misread jokes, thin pages, or outdated snippets and convert them into confident public advice. Read more Fake References What Fake AI References Do to Research Fabricated references can pollute scholarly writing, misallocate credit, and make evidence trails harder to trust. Read more Bad Citations When AI Citations Look Real but Fail AI citations can be fake, mismatched, or real links that do not support the sentence they appear to prove. Read more Legal Cases When Fake AI Case Law Reaches Court Fake legal cases show how fluent authority can become costly when professionals skip source verification. Read more Specialist Tools Why Expert AI Still Gets Things Wrong Specialist AI systems can reduce some errors but still produce misleading answers when users mistake branding for reliability. Read more Health Claims How to Slow Down Health Claims Health claims can feel urgent and personal, so they need careful checks of evidence type, source incentives, and medical context. 6 pages +Show subtopics Read more Accuracy Checks The pause that stops false health posts A brief accuracy check can interrupt the impulse to share dramatic health claims before checking their evidence. Read more Cancer Claims When alternative cancer cures replace care Cancer cure claims become most dangerous when they persuade patients to replace proven treatment with untested products. Read more AI Doctor Deepfakes When fake doctors sell health certainty AI doctor deepfakes exploit visual trust by making weak or fabricated claims look medically endorsed. Read more Hidden Cures Why hidden cures need stronger evidence Hidden-cure posts often turn weak clues into urgent treatment advice before human evidence can support the leap. Read more Natural Risks Why natural does not mean safe Natural supplements can still be active drugs in practice, especially when they interact with prescriptions. Read more Influencers When Trust in People Replaces Evidence People often judge claims through social identity and familiar messengers, which can blur the line between trust and evidence. 7 pages +Show subtopics Read more AI Impersonation Can AI Fake Trustworthy Messengers? AI tools can imitate polish, intimacy and local detail, making fabricated messengers or claims look more credible. Read more Shared Identity When 'People Like US' Becomes Proof A claim can feel safer than it is when it comes from someone who seems to understand your group or experience. Read more Local Groups When Local Trust Replaces Local Reporting In places with weaker local journalism, familiar community voices can become trusted shortcuts for rumours and public claims. Read more News Influencers When Personalities Become the News Source News influencers can make current affairs feel more accessible while shifting trust from editorial process to personality. Read more Parasocial Trust Why Familiar Creators Feel So Trustworthy Familiar creators can make public claims feel like advice from someone the audience already knows and likes. Read more Social Proof Why Popular Posts Can Feel True Visible approval signals can make weak or misleading posts look more reliable before readers check the evidence. Read more Lateral Reading Why Better Readers Leave the Page Good online verification often starts by leaving the page and checking what independent sources say about the claim. 7 pages +Show subtopics Read more SIFT Routine A Four Move Check Before You Share SIFT gives readers a simple sequence for pausing, checking the source, finding better coverage and tracing evidence. Read more Sponsored Content Can You Tell News From Sponsored Content? Native advertising can blur the boundary between news and promotion unless readers check labels and source context. Read more Video Checks Is This Viral Video From Here and Now? A video can be real but still mislead if it is old, clipped, or filmed somewhere else. Read more Surface Trust When a Website Looks More Reliable Than It Is Professional design, dot-org domains and confident biographies can create trust before the source has earned it. Read more Real Source Who Actually Made This Claim? A viral post is often only the last stop in a longer source trail, and finding the origin can change how much trust it deserves. Read more Fact Checker Study Why Fact Checkers Read Less First Professional fact-checkers often judged unfamiliar websites faster because they left the page sooner. Read more Missing Context What Did the Post Leave Out? A claim can be technically true but misleading when it hides timing, location, definitions, baselines, or comparison groups. 7 pages +Show subtopics Read more Inner City How One Baseline Bent a Gun Violence Claim The '95%' inner-city claim shows how county baselines and undefined terms can reshape a public-safety statistic. Read more Denominators The Missing Number Behind Viral Percentages A percentage can sound precise while hiding the population, period, or baseline that makes it meaningful. Read more Baselines What 'Up' Leaves Out of the Story Words like 'surge', 'record', and 'highest since' only make sense when the starting point is visible. Read more Pension Claim When a Bad Definition Drives the Whole Claim The viral pension-pot claim shows how a wrong definition can make a calculation feel morally obvious but factually wrong. Read more Old Footage When Old Footage Pretends to Be Breaking News Real footage can mislead when a new caption strips away the original date, place, and event. Read more AI Caveats Why Fluent AI Summaries Still Need Dates AI-written explanations can sound fluent while losing the dates, limitations, and uncertainty that keep a claim honest. Read more Mixed Feeds Why News Feels Harder to Sort When journalism, entertainment, activism, advertising, and synthetic content appear together, readers need stronger sorting habits. 7 pages +Show subtopics Read more Synthetic media Can you trust what looks real? AI images and other synthetic posts make fast feed judgments harder because realism, humour and persuasion can arrive together. Read more News or opinion Is this reporting or just a reaction? A feed post may look news-like, but its trust value changes depending on whether it reports facts, argues a view, or reacts to someone else. Read more Sponsored posts When an ad does not look like an ad Advertising inside ordinary posts can make persuasion look like recommendation, evidence or personal experience. Read more Influencers When influencers become accidental news sources Many people get news from influencers because it appears in entertainment spaces, not because they deliberately chose that source. Read more Platform cues Why popularity can look like proof Verification badges, view counts, avatars and polished video formats can feel like trust signals even when they only show visibility. Read more Source memory Why you forget where a story came from When stories arrive through search or social feeds, readers often lose the source cues that help them decide what to trust. Read more Name Claim What Exactly Is Being Claimed? Fact-checking becomes easier when the reader first states the exact claim instead of reacting to the whole post. 7 pages +Show subtopics Read more Fact Checkers How fact checkers decide what to test Fact-checkers first define the exact assertion of public interest so evidence can be applied consistently and fairly. Read more Vague Posts How vague posts become checkable claims A vague post becomes easier to test when the reader rewrites it as one precise question that evidence could answer. Read more Photos Videos What exactly does this image claim? Images and clips need claims about place, date, source, and context before reverse searches or archives can help. Read more False Causes When true numbers point to false causes A number may be accurate while the post around it pushes readers toward a cause the evidence does not support. Read more AI Claims Why fluent AI answers still need claim naming Fluent AI text can sound organised while blending facts, invented details, and vague assertions that must be split apart. Read more Search Wording Why the first fact check search fails Search tools may miss a fact check until the claim is rephrased with clearer names, dates, and technical terms. Read more News Deserts When Local Feeds Replace Local News Local groups can become fertile ground for false authority messages when reliable local journalism is weak or absent. 7 pages +Show subtopics Read more Council Checks How To Check A Council Rumor A viral claim about a council decision is easier to test when readers know which records, meetings and named officials to check. Read more SMF Dataset The Local Misinformation Numbers That Matter The SMF analysis shows why news-related posts in local feeds can be much riskier than overall group activity suggests. Read more Fake Authority When A Local Warning Looks Official False credibility often comes from borrowed institutional language, old images, manipulated screenshots or accounts that sound official. Read more Search Gaps When Searching Does Not Settle It Searching is useful but often weak for local claims because the record may be thin, duplicated or dominated by the original rumor. Read more Election Rumors Why Local Election Rumors Spread Fast Local elections sharpen misinformation because claims about candidates, services and identity can spread before reporters or officials respond. Read more Local Groups Why Local Rumors Feel So Believable Neighborhood posts can feel trustworthy because they name familiar places, people and problems, even when the evidence is weak. Read more Not Cynicism Sceptical Without Becoming Cynical Healthy scepticism asks for evidence and context, while cynicism treats everything as fake and can become another shortcut. 7 pages +Show subtopics Read more Trust calibration Trust less is not enough Teaching people to distrust everything can backfire unless it also helps them separate stronger evidence from weaker claims. Read more Blanket distrust When distrust becomes its own trap Cynicism can feel protective online, but it often replaces evidence with emotional shortcuts and familiar voices. Read more Deepfake doubt When fake images make real ones suspect Fear of synthetic media can help people spot fakes, but it can also give bad actors a reason to deny authentic evidence. Read more Repetition When familiar claims start feeling true Repeated claims can start to feel true, so cynical readers still need to check whether many posts share one real source. Read more Corrections When updates look like cover ups Responsible updates can look suspicious to cynical readers, even when changing the record is part of getting closer to truth. Read more Popularity cues Why viral numbers are not evidence Likes, views and confident delivery can make weak claims look credible even when creators have not checked the facts. Read more Own Research When Research Becomes Cherry Picking Responsible research means using a method for weighing evidence, not collecting links that confirm a preferred conclusion. 7 pages +Show subtopics Read more Loaded Search Are Your Search Terms Choosing the Answer? The words used to search a claim can quietly steer the results before any evidence is weighed. Read more AI Citations Can You Trust the AI Footnotes? AI can organize research, but fluent answers and realistic citations still need to be checked against original sources. Read more Lateral Reading Check the Source Before the Story Leaving the page to check who is behind it often reveals more than reading the page itself. Read more Opposing Evidence What Would Make This Claim Weaker? A claim becomes stronger when it survives the best counterevidence, not when a reader only finds supporting links. Read more Search Backfire When Googling a Claim Makes It Stick Searching to check a false claim can backfire when weak results surround the reader with repetition instead of verification. Read more Data Voids Why the First Results Can Mislead Obscure phrases and breaking claims can create empty search spaces where low-quality sources look more important than they are. Read more Platform Incentives How Feeds Reward Reaction Before Reflection Likes, comments, stitches, reposts, and algorithmic boosts can make performance feel more rewarding than accuracy. 7 pages +Show subtopics Read more Late Corrections Can corrections catch a viral post? Corrective notes may reduce engagement once attached, but their impact depends heavily on whether they arrive before a post peaks. Read more Accuracy Prompts Can one prompt make sharing more careful? A small reminder to think about accuracy can make people more selective about what they repost next. Read more Truth Rewards What if platforms rewarded accuracy instead? Changing social rewards so they favour true information can improve discernment without assuming users are indifferent to truth. Read more Like Counts When popularity starts to look like proof Visible popularity signals can make a claim feel socially validated before anyone has checked the evidence behind it. Read more Outrage Ranking Why angry posts keep climbing feeds Engagement-based feeds can turn anger, disbelief and correction into signals that push posts to more people. Read more Fast Formats Why screenshots outrun the source Low-friction formats can detach claims from sources, dates and missing context while making reposting feel effortless. Read more Primary Sources When Should You Go to the Source? Primary documents, regulator notices, court filings, and datasets can clarify what a viral interpretation leaves out. 7 pages +Show subtopics Read more Court Records Did The Court Really Rule That? A filing can prove that an argument was made, but only the court record shows whether a judge accepted it. Read more Robocall Trail How An AI Robocall Became An Evidence Trail The New Hampshire AI robocall case shows why official records matter for who did what, when and under which law. Read more Regulator Status Was It Banned Or Just Proposed? Many viral posts collapse warnings, proposals, settlements and final orders into one misleading word: banned. Read more Archived Pages What Did The Page Say Before? Archived pages can show what a source said before edits, deletions or quiet wording changes reshaped the claim. Read more VAERS Claims What VAERS Can And Cannot Prove VAERS can help detect possible safety signals, but raw reports cannot prove that a vaccine caused an event. Read more Public Numbers Where Did That Government Number Come From? A public number is harder to trust when the table, date, definition or method behind it is missing. Read more Ranking Why Did This Claim Find You? Critical thinking online starts with asking why a post reached you, not only whether the post itself is true. 7 pages +Show subtopics Read more Feed Controls Can You Really Train Your Feed? Buttons such as Not interested, hide, report and history controls can help, but their power varies by platform and feed surface. Read more Claim Route How Did This Claim Reach You? The same claim means something different when it arrives through search, a friend, an advert, a trend or a recommendation. Read more Viral Proof Viral Does Not Mean Verified A post can spread because it is useful, funny, shocking or divisive, so reach should never be treated as proof. Read more Outrage Signals When Outrage Looks Like Interest Watching, pausing, commenting or sharing can teach a system that a topic matters, even when your real reaction is rejection. Read more Similar Users When People Like You Shape Your Feed Recommendation systems can use patterns from people like you, making a narrow cluster of behaviour look like a wider reality. Read more Feeds vs Search Why Your Feed Is Not a Search Result Feeds often begin with a platform's prediction, not your question, which changes how a claim should be judged. Read more Screenshots Why Screenshots Make Claims Harder to Trust Screenshots and reposted claims often hide the original source, date, edit history, and context needed to judge them. 7 pages +Show subtopics Read more Fake Posts Did They Really Post That? A convincing screenshot can name a real person or institution while showing words they never published. Read more Late Notes What Corrections Cannot Catch in Time Corrections can reduce misinformation on live posts, but screenshots may keep the earlier false version moving elsewhere. Read more Old Images When Real Images Get New Stories A real image can still mislead when a new caption detaches it from its original time, place or cause. Read more Lost Satire When Satire Screenshots Stop Looking Funny A joke or fake-post account can become misinformation when a repost strips away the cues that marked it as satire. Read more Fake Authority When Screenshots Borrow Official Authority Screenshots can borrow institutional authority by making false claims look like official statements during fast-moving events. Read more Local Groups Why Local Screenshots Feel So Trustworthy Neighbourhood reposts can make weak evidence feel stronger because the messenger is familiar and the claim feels close to home. Read more Source Tracing Can the AI Answer Be Traced? A useful AI answer is only as strong as the sources, data, and context a reader can trace behind it. 7 pages +Show subtopics Read more Link overload Can more AI citations make answers worse? Long link lists can create a false sense of certainty when the right sources are not matched to the right claims. Read more Citation checks Does the cited source actually exist? A citation only helps when the reader can open it, identify it, and confirm that it is not invented or dead. Read more Wrong support When a real source does not prove the claim A working link can still mislead when it supports the general subject but not the specific sentence beside it. Read more Academic References When scholarly looking references are fake Fabricated papers and bibliographic details can make AI summaries look scholarly while hiding an unverifiable evidence trail. Read more News checks Why AI news answers need a date check AI news summaries can sound current while misidentifying sources, missing dates, or repeating outdated information. Read more Court cases Why fake AI citations reached the courtroom Court filings show why fake AI citations are not harmless formatting errors when professional duties depend on verification. Read more Synthetic Images When Seeing Is No Longer Enough Realistic AI images make visual plausibility a weaker shortcut for deciding whether an event, quote, or document is genuine. 7 pages +Show subtopics Read more Pentagon Fake How one fake explosion image moved attention fast The fake Pentagon explosion showed how a synthetic image can matter during the short window before corroboration catches up. Read more Source Chain What proves a viral image came from reality The strongest question is not whether an image looks real, but what source trail connects it to a real place, time and witness. Read more Reverse Search When reverse image search helps and fails Reverse image search can expose recycled photos, but newly generated images may leave no earlier copy to find. Read more Realism Trap Why fake images feel true at first glance Photorealistic AI images can make false claims feel verified before viewers ask where the image came from. Read more Inspection Limits Why spotting weird hands is not enough Hands, shadows and text can still reveal problems, but their absence no longer makes an image trustworthy. Read more Pope Puffer Why the Pope puffer fooled so many people The Pope puffer image showed that low-stakes, funny fakes can still train people to trust style over source. Read more Search pages Search topic, branch, or keyword... Scope 175 pages Search pages Clear
