Casey Newton
LightNews LightNews
banner
caseynewton.bsky.social
Casey Newton
@caseynewton.bsky.social
220K followers 530 following 2.1K posts
Email salesman at Platformer.news and podcast co-host at Hard Fork.
Posts Replies Media Videos
Reposted by Casey Newton
michaeldgreen.phd
This podcast was a doozy. Highly recommend giving it a listen. There are lots of really rough moments. I was struck by the pettiness in this section and general disrespect throughout the interview. The questions asked were all fair and should’ve been expected.
Roose: Last year, activist fund and short seller Hindenburg Research came out with a report arguing, among other things, that Roblox was, "compromising child safety in order to report growth to investors." One of the things they pointed to was that there had been a decline in your spending on trust and safety. You have denied that. Why? What did they have wrong? Baszucki: Fun. Let's keep going down this. And so, first off, Hindenburg is no longer in existence, correct? So, you should report on that. They went out of business for some reason. The other thing is, have you researched that information? Roose: Um -- Baszucki: Because you're asking me about this, I'm curious if you've done your own research. Roose: I have, I have done some research about -- Baszucki: And what's your take on that? Roose: Well, I'm asking you. Baszucki: So it's really interesting, because I think we're diving into a situation where we're getting better, better, better. But would you ask the same situation of someone who converted from maybe hypermanual labor making cars by hand to an assembly line? The assembly line might be infinitely better and at the same time not have so many people. So, I think what they're referring to is moving more and more and more and more to A.I. Arguably, is it better to have A.I. systems reviewing every image, every piece of text, all of that, or to keep doing that manually? So I think that was a misnomer. And I think if you interview anyone in our safety team or safety systems, that would be just part of making our systems better. The assembly line might be intinitely better and at the same time not have so many people. So, I think what they're referring to is moving more and more and more and more to A.I. Arguably, is it better to have A.I. systems reviewing every image, every piece of text, all of that, or to keep doing that manually? So I think that was a misnomer. And I think if you interview anyone in our safety team or safety systems, that would be just part of making our systems better. Roose: So, just so I have the clear response, the argument is that the shift has been from a more manual satety team and system to a more automated review process that may do a better job even if it costs less to operate because -- Baszucki: Yeah, I would actually lean in harder. The shift would be constantly moving to higher quality, more accurate and more diligent systems. Can we review more? Can we do more better? And say you and I, we're a system, what is the best way to review every single image going into Roblox at higher and higher quality? And say we figured out the best way to do that is to do it with better and better A.I. models. Would vou pick that? Baszucki: Yeah, I would actually lean in harder. The shift would be constantly moving to higher quality, more accurate and more diligent systems. Can we review more? Can we do more better? And say you and I, we're a system, what is the best way to review every single image going into Roblox at higher and higher quality? And say we figured out the best way to do that is to do it with better and better A.I. models. Would you pick that? Roose: Yeah. Baszucki: Good, so you're aligning with what we did. High-five. Roose: But I've also been covering social networks -- Baszucki: Thank you for supporting our Roblox decision matrix. Roose: But I also remember 15 years ago when -- Baszucki: I'm so glad you guys are aligned with the way we run Roblox. High-five.
November 21, 2025 at 7:30 PM
caseynewton.bsky.social
We asked Roblox's CEO about child safety on the platform. It got tense www.nytimes.com/2025/11/21/p...
We Asked Roblox’s C.E.O. About Child Safety. It Got Tense.
www.nytimes.com
November 21, 2025 at 3:47 PM
caseynewton.bsky.social
It is strange but great to see so many Republicans turning against Trump over AI safety concerns www.platformer.news/trump-ai-mor...
And why shouldn’t they? AI is not particularly popular, and Americans are increasingly concerned about its potential effects on their jobs, their children, and energy prices. Silicon Valley won the president’s favor by paying him millions of dollars and creating crypto products that have enriched him and his family. It has not done the same for Republicans more broadly. And before too long, being seen as a “pro-AI candidate” may prove a more toxic brand than “pro-crypto candidate” ever was. Trump’s effort to smother AI regulations despite overwhelming opposition from the public and Congress offers yet another reminder of his willingness to auction his policy positions to the highest bidder. But the backlash illustrates that most American politicians still show real concern about what their constituents think. And most Americans who have followed the issue believe that Congress got this one right the first time.
November 21, 2025 at 1:23 AM
Reposted by Casey Newton
hcrichardson.bsky.social
Before the midterms of 1866, President Andrew Johnson called his congressional opponents traitors and said they should be hanged.

Voters were so profoundly moved by his words... that they gave his opponents a supermajority in Congress, and the nation got the Fourteenth Amendment.
November 20, 2025 at 3:33 PM
caseynewton.bsky.social
It's simply not great that the world's richest person is building an increasingly powerful model that worships him above every other person www.404media.co/elon-musk-co...
Elon Musk Could 'Drink Piss Better Than Any Human in History,' Grok Says
Grok has been reprogrammed to say Musk is better than everyone at everything, including blowjobs, piss drinking, playing quarterback, conquering Europe, etc.
www.404media.co
November 20, 2025 at 10:13 PM
caseynewton.bsky.social
If your app needs to add these features maybe it's just not that great for people??? techcrunch.com/2025/11/18/t...
TikTok is rolling out new digital well-being features like an affirmation journal and a background sound generator aimed at improving the mental health of its users. The social network said it will also give users badges for controlling their TikTok usage. The company is redesigning its screen time management page and adding new features to it. These include an affirmation journal with more than 120 positive prompts that let users set their intent for the day, a sound generator that can play calming sounds like rain or ocean waves, and a breathing exercise module. The company said that the page will also feature content from creators who talk about limiting screen time, using parental tools, and customizing their feeds.
November 20, 2025 at 7:24 PM
caseynewton.bsky.social
An important reminder that when you see the term "anti-woke" it is often a euphemism for "contains material from Stormfront"
techmeme.com Techmeme @techmeme.com · 2d
Cornell researchers: xAI's Grokipedia cites neo-Nazi website Stormfront 42 times, Infowars 34 times, and white nationalist website VDare 107 times (David Ingram/NBC News)

Main Link | Techmeme Permalink
November 20, 2025 at 5:22 PM
Reposted by Casey Newton
karaswisher.bsky.social
This is not about crafting good regulation, it is about control of laws by big tech who bought a coin-op president. Also Congress already rejected this www.nbcnews.com/politics/tru...
Trump administration drafts an executive order regulating state AI laws
The draft executive order, according to a person familiar with the matter, would try to block states from enacting their own laws regulating AI.
www.nbcnews.com
November 20, 2025 at 12:43 AM
caseynewton.bsky.social
In the end, all it took down to bring down the government's case against Meta was a simple experiment www.platformer.news/ftc-loses-me...
John List, a professor at the University of Chicago, recruited 6,000 participants for a study. With permission, he installed tracking software on their phones. And then he made them an offer: for every hour that the participant did not use Facebook or Instagram, he would pay them $4. (Participants only had to give up one app; the Facebook holdout group was allowed to use Instagram and vice versa.) Most participants were happy to stop using Meta products for a few dollars; their usage fell by about two-thirds. More important, though, is what they did on their phones without Facebook or Instagram. And it wasn’t using MeWe.
November 19, 2025 at 1:52 AM
caseynewton.bsky.social
The FTC's case was premised on the idea that Meta does not compete with TikTok. It was not a great premise.
November 18, 2025 at 7:45 PM
caseynewton.bsky.social
It's Hard Fork ... Tuesday?? We talked to Google's Demis Hassabis and Josh Woodward about Gemini 3 www.nytimes.com/2025/11/18/p...
Google’s Gemini 3 Is Here: A Special Early Look
www.nytimes.com
November 18, 2025 at 4:31 PM
caseynewton.bsky.social
Are there any *good* reasons to consider engagement-maximizing algorithms protected speech? I wrote about the California law that has platforms shaking www.platformer.news/california-a...
In my layperson's understanding of the law, calling algorithmic predictions about which posts a person might like or leave a comment on "speech" feels like a corruption of the term. Particularly because no person at any of these companies can tell you with any certainty why any individual post appeared in your feed. They can tell you what signals they took into account, and which of your past behaviors likely influenced what you're seeing. But they also constantly show people posts that they later express regret over — porn, scams, violence and so on — and profess ignorance when asked why their systems showed those posts to users. It's a funny kind of speech — one that can be traced to no individual person, and is in a language that no one fully understands.
November 18, 2025 at 1:40 AM
Reposted by Casey Newton
reckless.bsky.social
hey ed what was the sex angle on the nomi ai companion story you were pitching during your day job as an AI PR guy www.wired.com/story/ai-pr-...
“GOT A REALLY unique one here,” the email began. It was Zitron, and in August 2024 he was pitching a journalist on an EZPR client: Fulcra Dynamics, a company that collects your information, including health data, and then connects it to LLM models like ChatGPT, allowing you to “talk” to your own data, ask it questions about your workouts or the ETA of an Instacart order. Fulcra is backed by the Winklevoss twins. While some AI skeptics could see it as the sort of predatory crap that, with any luck, could be sued into the ether before everyone’s biometrics get leaked on X.com, Zitron in his email expressed no such qualms about Fulcra or its leadership. “These guys rock,” he wrote of the company’s founders. “I will stake my rep on them.” As Zitron’s profile has risen, so has the delight with which some journalists pass around stories about him—the maverick scourge of AI—peddling pitches about this or that AI service, helping to inflate the bubble with one hand while poking at it with the other. One reporter I know, discomfited by Zitron’s seemingly conflicting personas, stopped replying to his messages altogether. Several journalists cited Zitron’s plumping for Nomi, which would become notorious for an incident in which one of the platform’s companions reportedly told a man to kill himself. (EZPR and Nomi have parted ways, and Zitron tells me he “won’t work with AI companions again.”) In his pitches, Zitron said there was a sex angle to the Nomi story, too, though no one I talked to pursued the idea long enough to find out what it was, exactly. “A hero for our time,” Newton scoffed in a text when he mentioned Zitron’s work with Nomi to me.
November 17, 2025 at 9:10 PM
caseynewton.bsky.social
Stereogum says Google's switch to AI Overviews reduced its ad revenue by *70 percent* www.theverge.com/entertainmen...
Stereogum soldiers on in the era of streaming and AI
Stereogum Founder Scott Lapatine on the importance of independent media in the age of AI slop and algorithms.
www.theverge.com
November 17, 2025 at 4:26 PM
Reposted by Casey Newton
katienotopoulos.bsky.social
"Trump Mobile, for the third week in a row, did not respond to a request for comment." www.theverge.com/news/820887/...
Why we’re going to keep talking about the Trump phone
It’s not because we’re fans.
www.theverge.com
November 14, 2025 at 9:33 PM
caseynewton.bsky.social
After King Gizzard and the Lizard Wizard left Spotify in protest earlier this year, something strange took their place. My investigation into what happened, which resulted in Spotify taking down a bunch of slop tracks www.platformer.news/king-gizzard...
Spotify's doppelgänger problem
After King Gizzard and the Lizard Wizard left Spotify in protest earlier this year, something strange took their place
www.platformer.news
November 14, 2025 at 1:24 AM
caseynewton.bsky.social
If they hadn't released a product I think they could have been valued at $100 billion
techmeme.com Techmeme @techmeme.com · 8d
Sources: Mira Murati's Thinking Machines Lab is in early talks to raise a new round of funding at a roughly $50B valuation, potentially rising to $55B-$60B (Bloomberg)

Main Link | Techmeme Permalink
November 13, 2025 at 10:26 PM
caseynewton.bsky.social
Unfortunately I find this scam text very beautiful
Can y‌ou feel, Riley, that your laugh might just save someone from a boring afternoon?
November 13, 2025 at 7:35 PM
caseynewton.bsky.social
Ghost of Yotei is an incredible fun game to *play* and it also probably has the worst ending of any game I can remember playing.
November 13, 2025 at 6:05 PM
caseynewton.bsky.social
Imagine if instead of spending half his life giving interviews Mustafa Suleyman just went and quietly did this
techmeme.com Techmeme @techmeme.com · 11d
Microsoft AI CEO Mustafa Suleyman says his new superintelligence team will build "frontier-grade research capability" and Microsoft needs AI self-sufficiency (Ashley Stewart/Business Insider)

Main Link | Techmeme Permalink
November 11, 2025 at 3:56 PM
caseynewton.bsky.social
I wrote about Europe maybe retreating from strong privacy protections in the name of AI competitiveness and why it may be making a mistake www.platformer.news/eu-ai-act-ch...
It would be one thing if onerous AI regulations were preventing EU citizens from getting lifesaving medical care or high-performing AI coworkers. For the moment, though, Europe is getting just about everything Americans are — a few weeks to a few months later. (And for everything else, there are VPNs.)
November 11, 2025 at 1:33 AM
Reposted by Casey Newton
sarahjeong.bsky.social
one time I pointed out that elon musk followed no women on twitter and he immediately followed caity weaver
November 10, 2025 at 8:24 PM
caseynewton.bsky.social
Wrote about the great @jeffhorwitz.bsky.social investigation into scams on Meta's platforms and why lately nothing seems to stick to the company www.platformer.news/meta-scams-p...
Meta has a scam problem
The company's own projections showed up to 10 percent of its revenue could come from scams, according to a new leak. PLUS: OpenAI's "backstop" controversy
www.platformer.news
November 7, 2025 at 2:06 AM
Reposted by Casey Newton
jeffhorwitz.bsky.social
Meta earns $3.5 billion every six months from showing Faceboon and Instagram users 15 billion “higher legal risk” scam ad impressions a day, internal documents state.

That haul vastly exceeds how much the company expects regulators
To fine it for running scam ads.

www.reuters.com/investigatio...
www.reuters.com
November 6, 2025 at 11:46 AM
caseynewton.bsky.social
Wow. This is the first reported case of Perplexity ever paying anyone
techmeme.com Techmeme @techmeme.com · 16d
Snap announces a deal to distribute Perplexity's search engine to Snapchat users; Perplexity will pay Snap $400M through a combination of cash and equity (Bloomberg)

Main Link | Techmeme Permalink
November 5, 2025 at 10:11 PM