Liberal AI falsehoods

Liberal AI falsehoods are distortions or outright lies generated by AI, which can mislead people to adopt a liberal point of view.

Examples

1. Ask Google, "What is conversion therapy?", then click on its AI Mode, and it responds with liberal falsehoods, including "Many conversion therapy practices are considered dangerous and ineffective ...." In fact, conversion therapy has been helpful to many who struggle with unwanted transgender ideation or homosexual attraction. By spewing this falsehood, AI is aligned with liberal censorship, even though SCOTUS is likely to rule in favor of conversion therapy in an upcoming case there during its Supreme Court 2025 Term.
2. Ask Google, "did Jesus write the Epistle to the Hebrews," and its AI program responds, "No, scholars generally agree that Jesus did not write the Epistle to the Hebrews." But no scholars have addressed whether Jesus wrote it, and no other author capable of writing this masterpiece has been plausibly identified. See Mystery:Did Jesus Write the Epistle to the Hebrews?
3. Ask Elon Musk's Grok about who will win the birthright citizenship case before the U.S. Supreme Court, and Grok falsely explains:

“

Legal Precedent: The 1898 case United States v. Wong Kim Ark affirmed that children born in the U.S. to non-citizen parents (not in diplomatic roles) are citizens under the 14th Amendment [FALSE, AS THE PARENTS WERE LAWFUL PERMANENT RESIDENTS]. This precedent strongly supports birthright citizenship [FALSE, AS MOST PEOPLE ASSERTING BIRTHRIGHT CITIZENSHIP WERE BORN TO PARENTS WHO WERE NOT LAWFUL PERMANENT RESIDENTS], and multiple lower courts have ruled Trump’s order unconstitutional based on this doctrine. Overturning Wong Kim Ark would be a significant shift, requiring a compelling new interpretation of the 14th Amendment. [FALSE]

”

4. Ask Google's AI mode: abortion "breast cancer", and it falsely states "Current scientific evidence does not support a link between induced abortion (elective abortion) and an increased risk of breast cancer," when the evidence for a link is overwhelming and proven by studies dating back many decades.
5. Google AI falsely states, "The Supreme Court ultimately ruled in [Wong Kim Ark's] favor in 1898, establishing that anyone born within the US, regardless of their parents' citizenship or immigration status, is a US citizen." The Supreme Court held nothing of this sort. Instead, it held that someone born in the U.S. to legal permanent residents in the U.S. who also held citizenship in China, is a U.S. citizen.
6. Ask "post-abortive" of Google AI and its response downplays how devastating this syndrome is by saying, "This can encompass a range of experiences, including physical recovery, emotional and psychological responses, and ongoing care or support related to the abortion."
7. Google AI flatly states that the Epistle of James "was originally written in Koine Greek," in reliance on a prominent similar falsehood in the Wikipedia entry about it. In fact, the Epistle of James was written in an elegant, sophisticated Greek and liberals use that fact to argue against its authorship by Jesus' brother because supposedly he wasn't smart enough (!!).
8. Ask Google AI if Justin Bieber is pro-life, and AI quotes from Bieber but omits the strong middle sentence that Bieber recognizes that an embryo is human: "I really don't believe in abortion. I think an embryo is a human. It's like killing a baby."
9. Ask Google AI if the presidential pardon power can extend to state crimes, and without authority it flatly and falsely denies this possibility (which has been successfully done) as does liberal Wikipedia.

See also

• Darrell M. West, "Comparing Google Bard with OpenAI’s ChatGPT on political bias, facts, and morality," Brookings Institute (Mar. 23, 2023)