India is betting on frugal innovation and its vast tech talent to catch up in the global AI race. Officials and startups believe it can compete by building low-cost language models in Indian languages and AI tools for local problems.
www.ft.com/content/75ad...
Why is this important? Scammers stole over $1 trillion from victims globally in 2023 using ads like the ones highlighted in this report, among other unscrupulous methods.
Advertising represented most of Meta's $164.50 billion revenue in 2024. However, fraudulent ‘investment’ scheme ads imitating mainstream news websites like BBC and The Indian Express continue to slip through Meta's filters. www.logicallyfacts.com/en/analysis/...
😔
3. To address the demands arising from user growth, Bluesky expanded its moderation team to approximately 100 moderators and continues to hire additional staff.
Any plans to engage with external fact checkers and moderators @safety.bsky.app ?
2. Other areas of concern included Antisocial Behavior (reports of harassment, trolling, or intolerance), Spam (excessive mentions, replies, or repetitive content), unwanted sexual content, and Illegal or Urgent Issues.
A few takeaways from this report:
1. In 2024, users submitted 6.48 million reports to Bluesky's moderation service, with 1.20 million of those reports concerning "Misleading Content." This includes issues such as impersonation, misinformation, and false claims about identity or affiliations.
AI-generated image of boxes of coats for donation with a caption that reads “kicking off the new year in service of our community”
everything about this new Meta AI bot is so dystopian but for me the AI-generated images of nonexistent donated coats is up there
Understanding how people relate to information is crucial, especially within their cultural and social contexts. Placing human experience and narrative at the centre of our analysis of digital platforms may prove essential for addressing societal divisions in the years ahead.
This @politico.eu report confirms that misinfo mainly affects already-convinced audiences (confirmation bias), that powerful narratives often stem from established cultural beliefs rather than viral falsehoods, and that human elements like charisma and trust matter more than factual accuracy.
Countering disinformation: While post-2016 reactions focused heavily on "bad actors" and techn manipulation, eight years of research have revealed this was overly simplistic. We lost sight of the fact that online activity represents only a fraction of how people form beliefs and make decisions.
This whole thread about MacKenzie Scott’s grants (and criticism of them) v. other foundation funding is very good, and this point is central.
A US health boss's murder sparked a torrent of online misinformation and calls for violence against other executives, suggesting a failure of social media moderation that analysts fear could translate into real-world harms.
www.yahoo.com/news/whos-ne...
The Global Engagement Center, a US State Department unit established in 2016, shuttered on Monday at a time when officials and experts tracking propaganda have been warning of the risk of disinformation campaigns from US adversaries such as Russia and China.
www.france24.com/en/live-news...
His immortalised words:
“I do not believe that I have been a weak prime minister. I honestly believe that history will be kinder to me than the contemporary media or for that matter the opposition in parliament.”
Dr Manmohan Singh taught an entire generation of Indians to dream big.
The former prime minister’s death deprives the nation of sage counsel at a time when his reforms are being replaced by empty bluster.
www.bloomberg.com/opinion/arti...
My first @businessinsider.com story is up!
This US survey of thousands of Threads and Bluesky users showed that Bluesky skews heavily Democratic while only 34% of Threads users identified as either Democrat or Independent.
Tips about Meta? pdixit@insider.com or Signal at pranavdixit.04.
❌ Three fake screenshots mimicking Trump's Truth Social have appeared on Bluesky (and other platforms), posing as posts on a rift between the president-elect and Elon Musk.
All three of these posts are fake.
THREAD
1/x
"The BBC has spent months tracking down Tibetan sources whose family and friends were detained and beaten. BBC Verify has also examined satellite imagery and verified leaked videos which show mass protests and monks begging the authorities for mercy."
www.bbc.co.uk/news/article...
Legacy media *is* dying and has been for a while. That’s due to failing ad economic models, vampiric private equity funds, bad management, and a general unwillingness by most to pay for news.
If a TikToker who thinks he’s Gen Z Walter Cronkite wants to portray himself as the “future” then lol
lol dudes like these explain news aggregated from actual reporters, chalk up a few hundred thousand followers on social media platforms, and declare themselves “the future”.
imagine being so “independent” that you rely on the tech platforms for your distribution.
I did. It's a good read. We ( @logicallyfacts.com) had our own research, supported by a much broader sample, to rely on.
AI-driven mis/disinformation did not have the apocalyptic impact on elections in 2024 that many suggested at the start of the year. As the year ends, @logicallyfacts.com assesses the pertinent question: how did AI influence voters worldwide, if at all?
www.logicallyfacts.com/en/analysis/...
Honda and Nissan are preparing to start negotiations on a possible merger, which could ultimately be expanded to include Mitsubishi, Japan’s Nikkei reported
Had the editorial BS radar beeping when I first watched the report.
It stank from the moment it aired.
www.thedailybeast.com/cnn-investig...