Lightnews — Scholar-powered news

Lune_Lys

@lunelys.bsky.social

260 followers 380 following 270 posts

Freelance graphic designer, translator, proofreader, (recently) subber. She/her but whatever else you fancy works fine too.

Posts Media Videos Starter Packs

Pinned

Lune_Lys @lunelys.bsky.social · Oct 18

Hi!! For people who don't know, a few years ago I was able to take part in this video series organized by @xiranjayzhao.bsky.social
It's something I'm extremely proud of & forever thankful for, for allowing me to connect with amazing folks & reconnect with my identity.
youtu.be/Sw2QySeH_vY?...

Raya's Queerbaiting of Southeast Asians - The Importance of Cultural Context to Queerness

YouTube video by Xiran Jay Zhao

youtu.be

5 22 87

Reposted by Lune_Lys

Sweeney Boo @sweeneyboo.bsky.social · 5d

I got to do the exclusive kickstarter cover for WITCHING SEASON which is now live on kickstarter!

A 150+ pages horror comic anthology from @cerealpancake.bsky.social and Lifeline Comics 🔮 🌙 and a long list of incredible creators!

www.kickstarter.com/projects/phi...

2 18 69

Reposted by Lune_Lys

Katherine Lam @katherinelam.bsky.social · 5d

I'm an illustrator who has worked in editorial, publishing, advertising, animation, and games. Recently though I've been drawing mechanophilia in MS Paint 🤖💦 #PortfolioDay

21 410 1.6K

Reposted by Lune_Lys

Jes & Cin @jesncin.bsky.social · Dec 6

Got a postcard from a 4th grade reader who likes Lunar Boy 🥰

So I whipped up a postcard from space to send to Cat from Brooklyn ❤️

#Art

A postcard with an illustration of pink clouds. A photo of Indu making a heart shape with his hands is taped on the card. There's a stamp with a picture of the Eyesun spaceship on it.

5 14 89

Lune_Lys @lunelys.bsky.social · 45m

Hi, I'm Lune, and swearing on my honor as a French, this statement is the absolute truth, and so overcome was I by emotions that I also forgot to take pictures.

Reposted by Lune_Lys

Lumi @lumi-sco.bsky.social · 9h

Rhinocéros sur @blast-info.fr avec @usul2000.bsky.social
Des groupuscules néonazis jusqu'aux salons feutrés de la bourgeoisie, l'extrême droite est unie. Ils s'entraident, se fréquentent, se soutiennent.
Ya les journalistes qui le montrent et ceux qui accompagnent.
www.youtube.com/watch?v=4M8r...

🦏 DISSOLUTION, CENSURE : LES MÉDIAS PAVENT LA VOIE DU RN VERS L'ELYSÉE

YouTube video by BLAST, Le souffle de l'info

www.youtube.com

8 200 450

Reposted by Lune_Lys

skeleton hugs (rabbit dad to MooMoo and formerly Tiffany) @skeleton-hugs.bsky.social · 2d

Maxo relaxo

Black and white rabbit MooMoo splooting alllll the way out

4 18 170

Reposted by Lune_Lys

Les Jours @lesjours.fr · 2d

🔴 Info Les Jours. Un ex-candidat du RN, écarté pour ses propos pro-Poutine et ses délires sur une «guerre civile» en France, juge aujourd’hui les demandes d’asile à la Cour nationale du droit d’asile.  🧵

🔗 lesjours.fr/obsessio...

4 160 130

Reposted by Lune_Lys

Ariane Lavrilleux @arianelavrilleux.com · 2d

Malgré le cessez-le-feu, l'aide humanitaire arrive tjs au compte-goutte à Gaza

Et Israël peut toujours compter sur les livraisons 🇫🇷de composants pour ses armes

Avec @antoinehasday.bsky.social on révèle un nouveau colis pour drones Hermes prévu le 20/10

disclose.ngo/fr/article/v...

Vente d'armes : livraison imminente de matériel français vers Israël

Une semaine après la signature du cessez-le-feu entre Israël et le Hamas, Disclose révèle qu’un lot de matériel fabriqué par la société française Sermat doit être expédié en Israël, lundi 20 octobre. ...

disclose.ngo

2 110 130

Reposted by Lune_Lys

Greg Egan @gregegansf.bsky.social · 3d

“Well-meaning” people are using machine translation to write Wikipedia articles in languages that they don’t speak themselves, accelerating the degeneration in quality of the web corpus for several languages with relatively few native speakers.

www.technologyreview.com/2025/09/25/1...

How AI and Wikipedia have sent vulnerable languages into a doom spiral

Machine translators have made it easier than ever to create error-plagued Wikipedia articles in obscure languages. What happens when AI models get trained on junk pages?

www.technologyreview.com

6 230 390

Reposted by Lune_Lys

Edwy Plenel @edwyplenel.bsky.social · 5d

La vérité du projet de budget Lecornu pour 2026 : pour l’assurance maladie et les missions non régaliennes de l’État, il est aussi austère que celui proposé par Bayrou. Pis, il épargne considérablement les ultrariches et les grandes entreprises. www.mediapart.fr/journal/econ...

Ultrariches épargnés, austérité dans la santé : le budget Lecornu déçoit d’emblée

Pour ce qui concerne l’assurance maladie et les missions non régaliennes de l’État, le projet de budget Lecornu pour 2026 est aussi austère que celui proposé en juillet par François Bayrou. Pis, il é…

www.mediapart.fr

19 290 420

Reposted by Lune_Lys

Jes & Cin @jesncin.bsky.social · 5d

Kon El Superboy at a concert for @dcforgaza.bsky.social

He ranks this concert a tactile/telekinesis!! The highest rating there is.

#Art #DCcomics #Superboy #Artsky

Kon El superboy holds out a phone and takes a selfie against a crowd of people at a concert.

1 11 40

Reposted by Lune_Lys

Jes & Cin @jesncin.bsky.social · 7d

WAIT WHAT-!!

Legally Blonde on MTV pro shoot screencap. Elle Woods looks at the board where the Ignyte awards announces Lunar Boy team Jes and Cin Wibowo as winner of the Outstanding Comics Team category. She says "Whoa!"

Is that my name up on that list
Does someone know that I exist

7 8 48

Reposted by Lune_Lys

B A R S U K @barssuk.bsky.social · Jul 5

The battlefield 🗡

54 4K 14K

Lune_Lys @lunelys.bsky.social · 6d

Omg congratulations!!!! 🥳🥳🥳

Reposted by Lune_Lys

Jean-Charles Geslot @jcgeslot.bsky.social · 9d

14 370 1.5K

Reposted by Lune_Lys

Jes & Cin @jesncin.bsky.social · 8d

Old Aminah doodle I am still fond of. Local trans butch from Lunar Boy.

#Art #OC #OCsky #Artsky

Aminah leans back. She looks off to the side. Aminah is wearing a tank top, shorts and a flannel tied around her waist.

2 34 96

Reposted by Lune_Lys

Micah @rincewind.run · 9d

this is absolutely my favorite protest innovation of 2025

"my opponents are violent antifa extremist terrorists"

"your opponents are dressed as giant inflatable frogs"

solid fuck you to the pepe crowd too

The Oregonian @oregonian.com · 9d

Things are happening at Portland's ICE facility tonight.

Read more of our protest coverage here: www.oregonlive.com/crime/2025/1...

42 590 2.9K

Reposted by Lune_Lys

Adam Serwer @adamserwer.bsky.social · 10d

www.wweek.com/culture/2025...

Can you put into words what that purpose is? You and the frog have really become icons of this protest.

There’s been a nice overall whimsy that has shown up. The frog is amazing. We had a bunch of other inflatable people here last night, too. I saw people playing Twister the other day.

What they rely on is fear. So by coming out in an absurdist manner, it speaks to them, to some extent, that we’re actually not that afraid.

It also dismantles their narrative a little bit. When they try to describe this situation as “war-torn,” it becomes much harder to take them seriously when they have to post a video saying [U.S. Secretary of Homeland Security] Kristi Noem is up on the balcony staring over the Antifa Army and it’s, like, eight journalists and five protesters and one of them is in a chicken suit.

96 3.2K 11K

Reposted by Lune_Lys

Bebop, Rocky, and Floor's Human @misserinrenee.bsky.social · 10d

I hate having a pen. At least when they're free roaming, they can't emotionally manipulate me by pouting between the bars like they're innocent little prisoners...(See below)

9 22 140

Reposted by Lune_Lys

Dr. Damien P. Williams Wants A Better Sociotechnical Paradigm @wolvendamien.bsky.social · 13d

This is genuinely some bullshit on bsky's part. A reaction meme of charlie kirk gets a Black person a permanent suspension, but people who actively doxx and spread hate against marginalized people not only get to stay, but het followed by mods?

We see you, bsky, and it's pretty fucking disgusting.

Aveta👾🇭🇹 @aliafonzy.blacksky.app · 13d

Unfortunate update: Link reached out to BlueSky and got email back now on reason.

His appeal is denied. He will remain permanently banned on BlueSky’s side.

Here are the screenshots he just sent me as of 8:22 AM EST(my time)

A screenshot of an email from the Bluesky Moderation Team with the subject line “Bluesky Account Behavior” sent at 22:15. The email states that a Bluesky account controlled by the recipient, with the handle @spacelawshitpost.me, posted a reply with an image where the alt text reads: Charlie Kirk sitting in a white T-shirt that says “Freedom.” A negative consequence follows. The email indicates this post violates Community Guidelines regarding violent or threatening speech, specifically listing threats of violence or physical harm against individuals, groups, institutions, or the general public, and incitement of self-harm or suicide.

A screenshot of an email continuation from the Bluesky Moderation Team displaying a bulleted list of Community Guidelines violations. The first bullet point describes threats of violence or physical harm as including threats or encouragement of violence against individuals, groups, institutions, or the general public. The second bullet point defines incitement of self-harm or suicide as content that encourages self-harm or suicide. The third bullet point explains wishes of harm as expressions of a desire for harm directed at a specific person or group. Below the bullets, a paragraph states that users may not use Bluesky to break the law or cause harm to others and that all users must be treated with respect. The next paragraph informs the recipient that as a result of these violations, your account has been taken down. A closing paragraph thanks the user for doing their part to keep Bluesky a welcoming and empowering place for all users and instructs them to email moderation@blueskyweb.xyz if they have questions or wish to appeal the decision. The email ends with the signature Bluesky Moderation Team.

A screenshot of an email from the Bluesky Moderation Team sent at 08:11 with the subject line Re: Bluesky Account Behavior addressed To: Hide My Email. The email reads:

Hi there,

Thank you for reaching out to appeal the takedown of your account on Bluesky. We have carefully reviewed your case, including the information you provided and the activity that led to the suspension of your account.

After investigating, we have determined that your account activity was in violation of our community guidelines, specifically promoting hateful content or being disruptive to other users. We take these guidelines seriously to ensure a safe and respectful environment for all our users.

As a result, we regret to inform you that your appeal has been denied, and your account will remain suspended. This decision is final and cannot be reversed. We understand that this may be disappointing news, but we must prioritize the well-being of our community and maintain a consistent enforcement of our policies. We encourage you to review our community guidelines to better understand the expectations we have for user behavior on our platform. Please do not create new accounts, since those will be taken down for ban evasion.

Thank you for your understanding and cooperation.

15 560 1.4K

Reposted by Lune_Lys

Catsuka @catsuka.bsky.social · 13d

Rebecca Sugar (Steven Universe) to direct a "Moomins" animated feature film.
variety.com/2025/film/ne...

81 1.7K 6.9K

Reposted by Lune_Lys

Des Walsh @deswalsh.bsky.social · 16d

People will say “what can we do it’s inevitable we better embrace it” but the companies are running at these kind of profit margins or lack there of so the question is how long can they last or even be cheap enough for companies to chose to use over staff who will always do better meaningful work.

Ed Zitron @edzitron.com · 16d

OpenAI has only made $4.3 billion in the first half of 2025, meaning that it will likely miss its $13bn revenue projection. Worse still, its burnrate is abominable, with $6.7bn on R&D compute, $2.5bn on inference, and *$2bn* on sales and marketing.
www.wheresyoured.at/sora2-openai...

OpenAI Is On Track To Miss Its Revenue Projections - And Its Burnrate Is Worse Than Ever
So, The Information recently put out one of the most weirdly-worded pieces in history, starting with the title: OpenAI's First Half Results: $4.3 Billion in Sales, $2.5 Billion in Cash Burn.

Reading this title, you might be forgiven for thinking that OpenAI was profitable, when it actually appear to be finding new, innovative ways to lose money at a never-before-seen scale. In fact, when you read the piece, you get a sense of how bad things actually are.

Also, keep in mind that these are numbers for the six months beginning January 1 2025 through the end of July 2025.

$4.3 Billion In Revenue Should Be Cause For Concern - And Suggests User Churn (While Also Feeling A Little Suspicious)
When I put out my piece about how much OpenAI and Anthropic actually make, some believed me alarmist when I said that I believed it brought in $5.26 billion in revenue by the end of the first half of 2025, a number I estimated using its leaked revenue numbers from January and May and guessing at compound growth rates.

I believe that this means that OpenAI has had some weird months of revenue. OpenAI hit $5.5 billion annualized (which translates to a single month times 12 — so $458 million) revenue in December 2024 (which I estimate was similar to January 2025), $10 billion in May 2025 ($833 million), and then, according to CFO Sarah Friar, "[its] first $1 billion month" in July 2025, though I'll add she signed off on projections that involve OpenAI making more than NVIDIA in 2029.

Let's Talk About OpenAI's Leaked Costs
$2.5 Billion "Cash Burn" - AKA The Cost of Goods Sold
OpenAI deliberately — I assume it leaked these numbers included the phrase "$2.5 billion cash burn" as a means of suggesting that it was actually making money, but having read the piece, I believe this refers to the "cost of revenue," so the literal cost of providing the services to both their ChatGPT users and API customers.

This bit needs a bit of clarity. According to The Information, the “cost of revenue” includes the cash that OpenAI has paid out for compute for inference — and not, it seems, anything significant beyond that, like employee salaries, sales and marketing (which is an entirely different subcategory within the piece), and so on.

Now, AI boosters will likely say “wow, $2.5 billion burn on $4.3 billion in revenue, that’s profitability!”

Yeah man, if a company didn’t have expenses, OpenAI would be super profitable.

Sadly, OpenAI has lots of expenses.

$6.7 Billion in Research & Development Burn
OpenAI is playing extremely silly buggers with these numbers, but this was $6.7 billion out the door, accounting for 155.81% of its revenue.

And, as a reminder, $6.7bn is also nearly 40% of its cash-on-hand at the end of the period. We’ll talk about this more shortly.

I am guessing here, but I believe this number is entirely compute for both training models and post-training, the process where you try and "teach" a model what outputs are good (and bad) once it's been fully trained.

A peer of mine suggested this number might also include salaries. I don't think that's the case, as OpenAI has previously broken out salaries separately in leaks, but it also used to break out the revenue from API and ChatGPT subscriptions.

Nevertheless, number go down.

Operating Loss: $7.8 Billion - Net Loss $13.5 Billion
This is a weird one, made weirder by The Information's choice of wording.

An operating loss in this case is OpenAI's revenue minus their operating expenses, but not including things like stock-based compensation and legal fees.

Its (much higher) net loss includes more unspecified losses, and OpenAI said that more than half of these losses were "due to remeasurement of convertible interest rights."

OpenAI Spent More Than $2.60 To Make $1 In The First Half Of 2025
The specific weasel wording (in OpenAI's case) of these (alleged) cash flow statements reported by The Information suggests that things are pretty bad.

In 2024, OpenAI made about $3.7 billion, and it burned $5 billion in the process, working out to it spending $2.35 to make $1.

Depending on how you slice OpenAI's current numbers, things are worse on a sliding scale. Between the $2.5 billion COGS, $2 billion on sales and marketing, and $6.7 billion on R&D, OpenAI is currently spending $2.60 to make $1.

If you take its "net loss" number — which I add involves a number of non-cash expenses — OpenAI is currently spending $4.14 to make $1.

This company SUCKS!

4 67 210

Reposted by Lune_Lys

Jes & Cin @jesncin.bsky.social · 16d

Quote with your queer characters.

Lunar Boy has trans people of different generations, questioning characters along with bi + pan rep! ✨

#Booksky

Indu, Aminah, Noah, Dewi, Ismaya standing against the trans flag

Noah and Indu lock arms with Yuliana, they both hold up the questioning flag behind them

Aminah holding Alon. Aminah has the bi and trans flag painted on her cheeks while Along wears a pan colored flower bracelet

18 53

Reposted by Lune_Lys

Zachary Berger @zacharyberger.bsky.social · 17d

Reminder: they're STILL taking all of our work + we STILL need (especially Cali) signatures on the petition supporting Bill AB412 - the AI Copyright Transparency Act.

TLDR= this'll require transparency about what's in AI datasets!

Signing takes 1 min. Once done, PLEASE SHARE!! tinyurl.com/4uxbp963

3 210 280

Reposted by Lune_Lys

Aura/Moom @moomanibe.bsky.social · 17d

I don't give a shit so I _will_ name and shame here

Director, Gen AI for Games Director, Gen AI for Games with verification
Netflix · Los Angeles, CA (On-site)

Key Responsibilities

Define and lead the Gen AI strategy across Netflix Games, helping to shape foundational infrastructure to player-facing features.
Identify and evaluate opportunities to apply Gen AI across our portfolio and studios both in the creation of new experiences and improving our workflows and pipelines.
Engage with game teams to develop prototype features and AI-native gameplay mechanics, ensuring technical viability and product-market fit.
Partner with external researchers, developers, and companies pioneering work within the Gen AI
Stay on the forefront of AI research and emerging tools, separating hype from high-leverage opportunities that align with our development and player goals.
Collaborate with our engineering and platform teams to build robust solutions and scale core capabilities (e.g., model inference, fine-tuning pipelines, data infrastructure, etc).
Build relationships with external developers, startups, and research groups at the forefront of Gen AI, identifying where their innovations can meaningfully accelerate or complement our internal efforts. Chart clear, strategic paths for partnership, integration, or co-development.
Influence broader company thinking around Gen AI, creative collaboration, and impact in games.

19 1K 2K