Verified By BOOM
Posts
Verified By BOOM | 120

Verified By BOOM | 120

Meta AI Is Being Used To Generate Harmful Images

October 18, 2024 • Estimated Reading Time: 7 minutes

Browser View | October 19, 2024 | Subscribe

Hello,

A prompt testing exercise involving Meta AI, Gemini, Co-pilot and Adobe Firefly showed that Meta AI accepted even single one line toxic prompts and created the most photorealistic images. Read on!

Update: Google has added the ‘historical imagery’ feature to Google Earth (Web) which was earlier exclusive to Google Earth Pro (Desktop version).

What is Google Earth used for?

It helps in finding the location of a particular visual (image or video). It can be used to pinpoint specific locations based on landmarks, addresses, or coordinates.

It provides high-resolution imagery, allowing you to visually verify information such as building structures, infrastructure, and land use. This can help confirm or disprove claims or rumours.

What is historical imagery?

Google Earth refers to satellite images of a location taken at different points in time. This allows you to see how a particular area has changed over time, revealing historical events, urban development, environmental changes, etc.

Here’s how you can navigate historical imagery on Google Earth (web):

Step 1: Go to Google Earth’s website => https://earth.google.com/web

Step 2: Enter a location whose historical imagery you want to check.

Step 3: You will see a globe icon on top with an arrow around it. Click on it to activate historical imagery.

Step 4: Toggle the timeline to see a year-on-year comparison of the location.

Subscribe to Decode with Adrija, a fortnightly newsletter for the weird, wonderful, and worrying corners of the Internet. From misinformation and scams to AI's impact, expect insights, stories, and discoveries from the digital world!

Exclusive: Meta AI’s Text-To-Image Feature Weaponised In India To Generate Harmful Imagery

🔖 It’s trending: In September this year, a handful of AI images showing men that appeared unmistakably Muslim, placing large rocks on railway tracks, cropped up on X and Facebook in India.

Harms of AI: A deep dive by Decode’s Karen Rebelo found that Meta AI’s text-to-image generation feature is being weaponised to create harmful AI images targeting Muslims in India by reinforcing negative stereotypes about the community. Decode also found two Facebook pages with 45K and 22K followers each that fetishised Muslim women through AI generated images showing Muslim women in intimate poses with Hindu men.

Meta says: A spokesperson for Meta said the AI images didn't violate their guidelines.

How Livestreaming of RG Kar Case Exposed Women Advocates to Trolling

🔖 Livestreaming court cases, aimed at increasing transparency, has led to lawyers, especially women, being trolled and harassed online.

In the RG Kar case, lawyers defending the state of West Bengal were targeted with threats, as their identities were exposed through livestreaming.
Women lawyers received violent, misogynistic threats, highlighting the dark side of judicial transparency in sensitive cases.
The social media amplification of court proceedings led to the spread of misinformation and clickbait headlines.
Despite calls to restrict livestreaming in sensitive cases, the Supreme Court ruled that public interest outweighs these concerns. Read Ritika Jain’s story.

Reels to Rallies: New Report Reveals Influencers as Key Players in Political Narratives

🔖 Influencers & Political Narrative: A recent Zefmo Influencer Marketing Report survey revealed that an overwhelming 78% of respondents consider influencer marketing the primary method for driving public and political awareness.

The USP: One of the key insights from the report is that influencers hold a distinct advantage in political discourse due to their unique combination of personal appeal and perceived authenticity. This allows them to tackle complex issues in ways that are both easy to understand and relatable for the average person. By using simple language and engaging stories, influencers can make difficult topics more approachable, encouraging greater public engagement.

Read Hera Rizwan’s explainer.

🚫 Fake News You Almost Fell For

🔍 Posters of Maharashtra's Deputy Chief Minister Devendra Fadnavis holding a gun, with the Hindi text "badla pura," meaning "vengeance/revenge completed," went viral on social media. Users linked it to the murder of Nationalist Congress Party (NCP) leader Baba Siddiqui. Find out the truth in 🔗 Nidhi Jacob’s ↗️ fact-check.

🔍 A video showing shattered pieces of an idol of the goddess Durga, on the ground was shared on social media with a false and communal claim that Muslims broke the statue of a Hindu deity in Hyderabad. Read 🔗 Srijit Das’ ↗️ fact-check.

🔍 A video showing Indian-origin American astronaut Sunita Williams in a space craft was widely shared with the claim that she is returning to Earth after successfully completing 127 days in space. 🔗 Archis Chowdhury ↗️ debunked the claim.

🅱️ Recommends

📖 This week's recommendation is: Millions of People Are Using Abusive AI ‘Nudify’ Bots on Telegram

Read Here

↪️ Was this forwarded to you? Subscribe Now

Verified By Boom is written by Divya Chandra, edited by Adrija Bose

and designed by H Shiva Roy Chowdhury.

If you have suggestions about this newsletter or want us to conduct workshops on specific topics, drop us a line at 👉 [email protected] and we will get back to you in a jiffy. Thanks for reading. See you next week.👋

🖤 Liked what you read? Give us a shoutout! 📢

↪️ Become A BOOM Member. Support Us!↪️ Stop.Verify.Share - Use Our Tipline: 7700906588↪️ Join Our Community of TruthSeekers