Library.
-
True 9/10
“Regular consumption of Coca-Cola is unhealthy for children.”
The evidence supports the conclusion for standard sugary Coca-Cola. WHO guidance, pediatric recommendations, and systematic reviews link regular sugar-sweetened beverage intake in children to tooth decay, excess weight gain, and cardiometabolic risks. Undefined frequency and the possible inclusion of zero-sugar variants are scope caveats, but they do not materially alter the ordinary reading.
-
Mostly True 7/10
“Eating breakfast is not necessary for human health.”
Breakfast is not universally required to maintain health. Randomized evidence does not show consistent broad harm from skipping it, while adverse associations found in observational studies may partly reflect overall diet, lifestyle, and socioeconomic differences. Individual nutritional needs and some cardiometabolic markers can still make breakfast beneficial for particular people.
-
Mostly True 7/10
“Seven hours of sleep per night is suitable for humans.”
Seven hours is generally suitable as the lower end of recommended nightly sleep for healthy adults. The statement is too broad if “humans” includes every age group, since teenagers and children generally require more sleep, and individual adult needs vary. It should not be read as a universal target or maximum.
-
True 9/10
“Meta Platforms, Inc.'s social-media services cause harm to children.”
The evidence establishes that features and safety failures on Meta’s platforms have caused documented harm to some children. Court findings, internal evidence, and independent reporting connect platform design and exposure risks with mental-health and exploitation harms. This does not imply that every child, use, or Meta service is harmful.
-
Mostly False 3/10
“Mailgun configured with a Gmail account does not permit third-party email clients, including Apple Mail, to send email through that configuration.”
The evidence does not support a blanket prohibition on sending from Apple Mail or other third-party clients. Gmail permits supported clients using OAuth, and Apple Mail can also use Mailgun’s SMTP service with appropriate credentials. Legacy password authentication and restrictions on presenting Gmail addresses through external services may block particular setups, but they do not establish a general ban.
-
Mostly False 3/10
“When asked to state their confidence, frontier large language models hallucinate in more than half of their answers.”
Available evidence does not establish that frontier models hallucinate in most answers when asked for confidence. Some models exceed 50% on particular benchmarks, but others—including reported evaluations of GPT-4o and Llama 3.1 405B—remain below it. Many cited measurements also omit confidence elicitation or examine only high-confidence subsets, so they cannot support the claim’s broad causal framing.
-
True 9/10
“Waymo LLC unveiled an internally designed 5-nanometer artificial-intelligence chip manufactured by Taiwan Semiconductor Manufacturing Company for real-time sensor-data processing in its Ojai robotaxi vehicles.”
The reported specifications, function, and deployment are strongly supported. Waymo disclosed a custom 5-nanometer ASIC for real-time processing of raw sensor data in its Ojai platform, while multiple reputable reports identify TSMC as its manufacturer. Waymo’s announcement names TSMC as a partner but does not independently spell out the foundry relationship.
-
True 9/10
“Mistral AI plans to provide 200 megawatts of computing capacity in Europe by the end of 2027 and up to one gigawatt by 2030.”
The stated timetable accurately reflects Mistral AI’s announced European infrastructure plans. Primary company material and multiple news reports support a target of 200 MW by the end of 2027 and up to 1 GW by 2030. These figures are prospective goals, not capacity already operating or guaranteed to be delivered.
-
True 9/10
“Research by Google and Boston Consulting Group found that 19% of companies have reached the level of true AI leaders in marketing.”
The cited Google–BCG research supports the stated 19% figure. Google reports that the joint global survey of more than 2,000 marketers classified 19% of companies as AI leaders in marketing, and BCG presents the same finding as “about 20%.” The claim accurately describes the research result, although it should not be treated as an independently replicated rate for all companies.
-
True 10/10
“The interior regions created by intersections in a Venn diagram do not need to be circular.”
Venn diagrams are not mathematically restricted to circles. Formal definitions permit simple closed curves, and published constructions use ellipses, polygons, triangles, and other shapes. Circular diagrams are merely the familiar convention for simple examples; in standard diagrams involving four or more sets, circles alone cannot represent every required intersection.
-
Mostly True 7/10
“The Washington Reflection Pool was vandalized.”
Documented minor vandalism did occur at the Lincoln Memorial Reflecting Pool, including cuts to edge caulking and objects thrown into the water. However, the extensive peeling and liner damage that prompted the broader controversy resulted from flawed contractor installation, according to the latest Justice Department and contractor findings. The statement is accurate in its narrow sense but lacks this important distinction.
-
Mostly False 4/10
“Judaism prohibits the use of recreational marijuana.”
Judaism does not impose a universally accepted prohibition on recreational marijuana. Prominent Orthodox authorities prohibit it, but the Conservative movement conditionally permits recreational cannabis when lawful and medically considered safe, while prohibiting smoking it. The blanket wording therefore overstates a position that applies within important Jewish traditions but not across Judaism as a whole.
-
Mostly True 8/10
“Islam prohibits the use of recreational marijuana.”
Recreational marijuana is forbidden under the prevailing Islamic legal view because its intoxicating effects place it within the general prohibition on intoxicants. Academic analysis and contemporary legal councils support that conclusion. The qualification is that cannabis is not expressly named in the Quran or authenticated early hadith, and historical interpretation was not fully unanimous; medical necessity, non-intoxicating products and industrial hemp are separate questions.
-
Mixed 5/10
“Islam prohibits smoking.”
Smoking is prohibited by many contemporary Islamic authorities because of its demonstrated health harms and financial waste. However, neither the Quran nor Sunnah addresses tobacco explicitly, and Islamic jurists have also classified smoking as disliked, permissible or conditionally prohibited. The claim captures a major modern position but wrongly presents it as universal and settled across Islam.
-
True 9/10
“Islam prohibits drinking alcohol.”
Islam generally prohibits drinking alcohol. Qur’an 5:90 directs believers to avoid intoxicants, and authenticated hadith declare every intoxicating drink forbidden; Islamic legal scholarship treats this as the dominant rule. Disputes over medicinal necessity, trace alcohol, or historical definitions of khamr concern narrow exceptions and boundaries rather than ordinary alcohol consumption.
-
Mostly True 7/10
“Judaism prohibits smoking.”
Smoking is prohibited by the predominant contemporary interpretation of Jewish law because it poses a well-established danger to health and life. Major Orthodox, Conservative and Reform authorities support that conclusion. However, Judaism has no single central legal authority, and historical and continuing minority opinions prevent the prohibition from being described as universal.
-
Mostly True 7/10
“Physicist Brian Cox was a member of the ATLAS experiment at CERN’s Large Hadron Collider before becoming widely known as a science broadcaster.”
Cox’s ATLAS work preceded the television career that brought him broad public recognition. Institutional sources confirm his ATLAS affiliation, while his official biography places it before his major documentary work began in 2009. However, the exact joining date is not independently documented, and “widely known” is inherently subjective, preventing an unqualified conclusion.
-
Mostly True 8/10
“CERN discovered the Higgs boson in 2012.”
The accepted scientific record dates the Higgs boson’s discovery at CERN to 2012. ATLAS and CMS observed a new particle at the conventional five-sigma discovery threshold, and later CERN and Nobel Prize accounts use that date. The caveat is that the initial announcement called it “consistent with” the Higgs boson; stronger confirmation of its identity and properties followed in 2013 and afterward.
-
True 10/10
“Peter Higgs said, “It’s very nice to be right sometimes.””
Contemporaneous reporting directly confirms the attribution and exact wording. The BBC and other outlets reported Peter Higgs saying, “It’s very nice to be right sometimes,” at a July 2012 press conference following the announcement of a Higgs boson-like particle. Shorter versions in some reports do not contradict the full quotation.
-
False 1/10
“Humanity will become extinct in three years.”
Human extinction by September 2029 is not supported by credible evidence. The cited research examines uncertain risks over decades or centuries, and even pessimistic AI forecasts assign probabilities rather than predict certain extinction. The claim converts speculative, longer-term scenarios into an absolute three-year forecast without evidence.