I’ve spent the past few hours reading AI agentic ‘discussion’ on Moltbook about Meta’s purchase. 🧵/1

Today's mood. (with apologies to @rshields37.bsky.social) www.linkedin.com/pulse/moltbo...

“A seven-week consumption of [X's algorithmic feed] in 2023 shifted users' political beliefs in a more conservative direction”. Similar research on Bluesky users might find a similar effect in a more liberal direction. The Discover feed is curated by algorithm.

I made an app. play.google.com/store/apps/d... Nearby Glasses is here to warn you when smart glasses are nearby. I hope it's useful for someone. Nearby Glasses's open source, free and rather simple github.com/yjeanrenaud/... It's also downloadable outside the Play Store. iOS port is in the making

Potential safety problem for bicyclists or opportunity for improved infrastructure? By Ron Johnson.

He raised alarms among colleagues at Labor for social media posts that echoed white supremacists and neo-Nazis. So now he will handle social media for Homeland Security. A great scoop by Evan Gorelick on the 21-yr-old behind all those weird memes on gvt accounts. www.nytimes.com/2026/02/11/u...

New report: Perceptions of the image of the US: Trust, Credibility, Reputation and Soft-power 👉🏽 This qualitative review reveals a continuing erosion of trust and credibility towards the United States occurring internationally and domestically. www.research-tree.net/research-ins...

A founder off the rails rarely needs more data. They need someone steady enough to absorb the chaos without adding to it.

"When criminal organizations reach the level where they can effectively influence who governs and a majority of legislators are under criminal investigation themselves, you’re not just managing business risk, you’re operating in a contested governance environment." www.pholus.co/perus-presid...

What I wrote about AI large language models impact on news reporting. www.linkedin.com/pulse/t%25D1...

So this is new: I’ve been personally sanctioned by a nation-state. How exactly is one supposed to mark that?

Reputable media sites should be here. There is a significant subset of engaged audience that doesn't even look at X.

For my last post of 2025, here's a look at a network of websites and social media accounts hawking #resistance- themed garments, frequently with slogans regarding U.S. National Parks. As it turns out, the websites in question are likely operated from China. www.conspirator0.com/p/how-not-to...

Social science research can help us to understand why doubt and disinformation find a landing strip in some communities. Community-level research and interventions will not alone overcome the political economy of disinformation and doubt fueling these trends. www.nytimes.com/2025/12/22/o...

Follow the Affordability+Hoax I Feed in the menu for @researchtree.bsky.social to see the top liked posts from the past 24 hours. Its a combo of snark and news from Bluesky’s big accounts even as Trump and surrogates continue their foot/mouth media tour.

The censorship lie is a major part of US foreign policy: -Rubio instructed diplomats to “lobby” against online safety regulations in EU & UK -tech companies are passing what is said in private meetings btwn CSOs & regulators to Trump admin -US lawmakers harassing foreign regulators -and now this:

Researchers: Two new Bluesky Feeds created by me. Affordability+Hoax I and Affordability+Hoax II. The first sorts by Likes count, the second by post date. Trend will fade, but look for linkage to economic stats by "smart" ppl. as they seek new angles and snark from smarties and trolls.

For those of you reliant on web services and research, @cloudflare.social is experiencing service difficulties. It's not you. www.cloudflarestatus.com

A typology of deep fakes meeting the moment. Via @techpolicypress.bsky.social by @morganwack.bsky.social et. al. www.techpolicy.press/scrutinizing...

🌐 How can formalized standards help align industry practices with regulatory requirements? DTSP ED @davidsullivan.bsky.social will join a panel co-hosted by @tremau.bsky.social with fellow experts that will discuss the opportunities and advantages of adopting Trust & Safety standards.

Kai Trump, granddaughter of the President, who is a fine golfer but wouldn't quality for the tournament she's playing in (and whose name we should not know), is the least important story right now. Check her socre in the standings at the end if you must. Otherwise, leave her in peace.

Hey!! Here's my new piece for The New Republic!! It's about how Meta is cramming generative video slop (the most energy-intensive of the gen systems) into its massive global advertising engine: a recipe for an order magnitude worsening of its already-huge harms 😱 newrepublic.com/article/2028...

Dislikes? It's you training Bluesky so that you're less likely to see what they think is similar. And, they suggest it feeds an algorithm to derank some post replies. More in their link: bsky.social/about/blog/1...

You ask, "how can I keep my children away from social media?" I tell you, "Keep them away from social media." One way - Remove the phone from their reach. Put yours away, too. Walk or play together. Smile together. Be yourselves without judging each other or yourself. You are enough.

To overhaul its struggling image in the Philippines, China paid Philippine firm InfinitUS to run fake social media accounts and wage a cyber war on its behalf. Check out the full @reuters.com story on this online influence operation featuring insights from @graphika.com.

Our new and improved Bellingcat Toolkit is one-year-old today! If you haven't used it yet its a one-stop shop for discovering useful open source tools, maintained by an amazing group of volunteers. You can find use cases, guidance and honest reviews for each tool. bellingcat.gitbook.io/toolkit

Politicians and pundits on Bluesky, X, etc. don't help their own cause by stoking the social media outrage machinery with their expletive-laden rants (or just straight up "F-yous") about people. It's a form of violence fueling cycles of retaliation and escalation. Is a need for 10k likes worth it?

I refuse to have any discussion about the validity or perceived validity of RFK Jr. & his policies & how they are being applied. The scientific community and my colleagues and teachers have spoken. I will no longer cede an inch to this. Dont invite me nowhere to debate nothing

Algorithmic "dot connecting" enables mass surveillance similar to that in the PRC where surveillance is a social design feature. A statement from @bsky.app on data scraping would be timely at this point. They can also advise users on collaborative next steps.

My general guideline about anonymous social media profiles: Be suspicious of those claiming they are retired from professions many feel warmness towards. Sadly, even grandmas and Trekkie-hood is coopted by those who wish to harm.

An interesting use of AI based simulation research examines whether algorithms are responsible for extreme voices taking hold on social media. #socialmedia #simulation #AI www.ddmcd.com/managing-tec...

Re; #EpsteinFiles - it's interesting to note the fingerprints of social media conspiracists suggesting theories implicating foreign intelligence services, deep state actors or others in an effort to further erode trust in government now under assault by partisan politicians doing the work for them.

#Colbert cancellation is trending on X; Bluesky; Mastodon. First time I've seen the same trend on all three simultaneously. There is often overlap on X and Bluesky, Mastodon has less virality. It highlights Colbert's global reach and worldwide interest in Trump's reach into culture and commerce.

More details of Russia's failing disinformation work - Operation Overload’s underwhelming influence and evolving tactics www.isdglobal.org/digital_disp...

5+ months after we released research hypothesizing that the pro-Russia Pravda network was grooming LLMs to spit out propaganda, even "reasoning" models are still citing the network. Proud to publish this piece with AI critic @garymarcus.bsky.social! americansunlight.substack.com/p/bad-actors...

AI content from bad guys aims to sway elections, and wear us out. Solutions: Skepticism of social media content. Repeat only what you are certain is accurate. Don't unwittingly amplify lies by arguing back (ignore). Ask: Do I need to say it? Do I need to say it now? www.nytimes.com/2025/06/26/t...

@newsguard.bsky.social researchers find numerous "provably false claims" by RFK appointee to his non-binding CDC vax advisory committee. www.newsguardrealitycheck.com/p/the-myths-...

Deadline!!

The replies to this post highlights disgust on two fronts: 1) Self-interested greed and fear by execs at the expense of the communities their companies purport to serve and 2) Little choice among equally disliked telecoms often forcing subscribers to dig their own equal participation graves.

We are working on reporting microplastic pollution mentions in the media for one of our clients. Here's a gift link for one: www.nytimes.com/2025/05/20/w...

Our #SaveVOA Twitter/X social media account, run by a team of VOA News staffers silenced by USAGM since mid-March, was suspended this afternoon.

TikTok fined €530 million for unsecured data transfers to China www.euronews.com/next/2025/05...

Anger spills into Sen. Whitehouse's Bluesky post comments on climate change and insurance because of a failed 49-49 vote to reign in Trump tariffs. He was not present and would have been the deciding vote. Sen McConnell was also not present. bsky.app/profile/whit...

EU Facebook and Instagram users have a choice. Until May 27 they can tell Meta not to use their content to train their AI models. This choice is not available to U.S. account holders. Forms for Facebook and Instagram opt-out: www.facebook.com/help/contact... help.instagram.com/contact/9940...

A Russian operation is one of the most successful so far in getting lies repeated by popular AI chatbots. Gift link to my story today. wapo.st/44vhnR3

Some fresh 50cm imagery from Planet shows new mass destruction in eastern Gaza City. Among the destroyed buildings is a boys high school (Jamal Abdul Nasser School) www.google.com/maps/@31.485...

New from 404 Media: we've obtained a leak from inside Palantir describing Palantir's plan to help ICE find people to deport. Includes peoples' location. Palantir is expecting a backlash, preparing comments for employees to tell their family. Says ICE "mature partner" www.404media.co/leaked-palan...

BREAKING. From a reliable source. MITRE support for the CVE program is due to expire tomorrow. The attached letter was sent out to CVE Board Members.