Trending
rescbr's Avatar

rescbr

@rescbr

159
Followers
736
Following
539
Posts
04.07.2023
Joined
Posts Following

Latest posts by rescbr @rescbr

E aqui, que um dos gatos não cobre o cocô? E depois que ele faz o serviço, fica miando para avisar todo mundo que a obra-prima está disponível para apreciação.

13.03.2026 14:14 👍 2 🔁 1 💬 1 📌 0
Video thumbnail

www.percepta.ai/blog/can-llm...

As a research lark at Percepta, Christos embedded a computer into an LLM, showed that it could solve the hardest Sudokus, and then as a side bonus built an exponentially faster attention

11.03.2026 21:44 👍 302 🔁 55 💬 19 📌 42

o fumódromo LGBT neurodivergente da internet

11.03.2026 17:05 👍 814 🔁 241 💬 9 📌 8

Para ver que este problema não está mais restrito ao lado de cá do globo.

11.03.2026 12:10 👍 0 🔁 0 💬 0 📌 0

I really don’t understand these guys. EVs usually have more horsepower and better acceleration compared to gas powered cars.

Why would any macho man prefer a stupid gas guzzler when it has worse performance?

10.03.2026 18:16 👍 5 🔁 0 💬 1 📌 0

Yes, hello, is that God? Yeah, hi. I have a complaint. You gave me this brain that’s designed for finding berries and avoiding lions and now people are ‘just circling back’ to see if we can ‘move the needle’ on ‘key initiatives’? NONE of those things are berries.

10.03.2026 11:07 👍 4544 🔁 1250 💬 46 📌 53

Contar pra vocês de uma conquista muito foda que rolou no trampo.

Em 2023 eu fui contratado como head of accessibility na Yape, o maior app de pagamentos do Peru, para iniciar o trabalho de tornar o bagulho acessível a leitor de tela.

09.03.2026 13:47 👍 155 🔁 41 💬 6 📌 14
Post image

If I had a nickel for every time Franz Ferdinand was involved in the precursor events of a world war, I’d have two nickels, which isn’t a lot, but it’s weird it happened twice

09.03.2026 08:34 👍 11746 🔁 2797 💬 67 📌 82
Post image

🙏✨

07.03.2026 20:39 👍 488 🔁 132 💬 3 📌 4
Preview
Anthropic just mapped out which jobs AI could potentially replace. A 'Great Recession for white-collar workers' is absolutely possible | Fortune The red jobs are what AI is doing now, Anthropic says. The blue ones are what it's coming for next.

fortune.com/2026/03/06/a...

07.03.2026 14:48 👍 2 🔁 2 💬 3 📌 0
Preview
The Invisible Siege: How Insurance Markets, Not Missiles, Closed the Strait of Hormuz And Why the Disruption Will Last Four to Sixteen Months Longer Than Any Model on Wall Street Currently Prices

Traffic won't resume soon.

shanakaanslemperera.substack.com/p/the-invisi...

06.03.2026 19:28 👍 10 🔁 3 💬 1 📌 1
Video thumbnail

⭕️Une #Iranienne documente le moment où les forces d'occupation #Américaines et #Israéliennes ont pris pour cible un quartier civil en #Iran.

05.03.2026 20:43 👍 20 🔁 13 💬 2 📌 3
Morocco solar farm

Morocco solar farm

We stole the sun from the sky

To power the Machines of our invention

And spite the Gods that abandoned us

20.10.2025 10:05 👍 240 🔁 29 💬 6 📌 8
Preview
What the World Needs Now is Forums Sweet Forums - Mal’s Musings and other thoughts from a forums purveyorI was seven years old when I got my first video game console. It was Christmas, and I actually got two! A GameBoy Po...

I'm finally ready to show off the thing I've been working on: atBB (pre-alpha)!

It's a forum platform built on top of atproto.

I wrote up some thoughts about why I am interested in building forums: pckt.blog/b/mal-s-musi...

1/2

25.02.2026 01:16 👍 34 🔁 14 💬 2 📌 0

Both showers and public transit rides have 3x multipliers to Thinking Time. You'd think stationary bikes would be too because they're both moving and mindless but actually they're 0.5x at best. Long flights have 10x to Thinking Time but only if you're in the sweet spot of tired but not TOO tired

05.03.2026 02:45 👍 648 🔁 111 💬 17 📌 9
Post image Post image Post image Post image

What you're looking at is not a sunrise — it's the Russian LNG tanker ARCTIC METAGAZ (IMO 9243148) struck by a massive explosion in the Mediterranean this morning. Photographed by crew aboard a merchant vessel, via Vanguard Tech.

03.03.2026 23:04 👍 420 🔁 133 💬 23 📌 35
Post image

Just one more week...

03.03.2026 13:49 👍 54 🔁 8 💬 3 📌 0

O agenciamento da extrema-direita de "reencantar" o mundo por meio da religião, resgatando o aspecto cósmico do cristianismo medieval, tem uma relação direta com a ressurgência da guerra santa em escala "fractal".

02.03.2026 16:42 👍 336 🔁 43 💬 6 📌 4
Preview
The old internet is gone but we can still save the new one A love letter to the internet that once was

“On the Internet, nobody knows you’re a dog.” Nobody knew you were a teenager either. And that liberation, that sense of playfulness and experimentation and reinvention that is so crucial to adolescent identity-shaping, got bulldozed by social media

www.newstatesman.com/politics/uk-...

03.03.2026 10:26 👍 15 🔁 6 💬 1 📌 0

Probably they saw what happened in Venezuela and want to seize the opportunity.

02.03.2026 04:02 👍 0 🔁 0 💬 0 📌 0
Post image

Sarandos’ first interview since they exited the race for Warner is pretty lit.

@bloomberg.com $NFLX
@lucasshaw.bsky.social
www.bloomberg.com/news/article...

01.03.2026 22:04 👍 752 🔁 128 💬 20 📌 23
Video thumbnail

Virtually a first person view of the Iranian drone hitting the high-rise building in Manamah, Bahrain.

28.02.2026 18:57 👍 336 🔁 72 💬 14 📌 12
Post image Post image

In 2015 (!), Tim Urban (waitbutwhy) published this post which included this graph.

waitbutwhy.com/2015/01/arti...

Bluesky's AI discourse affliction is that it still holds the distorted view in the second image.

I worry a lot about what's going to happen to society when we reach "The fuck??"

20.12.2025 18:27 👍 41 🔁 2 💬 6 📌 4
Post image

😭😂🤣😂😭

02.12.2025 19:35 👍 917 🔁 287 💬 4 📌 16

Daí joga na internet que pra cobrir a falta do imposto em bet vai ter que taxar 14 bi de tecnologia que daí vai ver o rebuliço nas redes. Cozinha isso por umas semanas. Faz teatrinho. Vai subir na opinião pública e ainda pega o dobro em $.

Mas sabemos quem paga o andar de cima. Pilantras.

26.02.2026 02:59 👍 1 🔁 0 💬 1 📌 0

Só ver que a taxação das bets saiu do PL Antifacção. 30 bilhões.

Teria que ter botado o pau na mesa e ameaçado o congresso pra botar o imposto nas bets de volta usando a opinião pública.

Ninguém vai ser contra imposto em bet.

26.02.2026 02:59 👍 1 🔁 0 💬 1 📌 0
A horizontal bar chart titled “Model Detection Breakdown (%)” with a subtitle explaining: “Each bar is continuous and split into Green, Amber, and Red, sorted by Green %.”

Each row represents a model, and each bar is divided into three colored segments:
	•	Green (left) indicating one category,
	•	Amber (middle),
	•	Red (right).

Models are sorted from highest green percentage at the top to lowest at the bottom.

At the top, models like:
	•	Claude Sonnet 4.6 — 94.9% green, 4% red
	•	Claude Opus 4.6 — 92.7% green, 5% red
	•	Claude Sonnet 4.6 (High) — 92.7% green, 5% red
	•	Claude Opus 4.5 (High) — 90.9% green, 9% red
	•	Claude Opus 4.6 (High) — 89.1% green, 7% amber, 4% red

These top models have large green bars and very small red segments.

Mid-tier entries include:
	•	Qwen3.5 39B A17b — 65.5% green, 20.0% amber, 14.5% red
	•	Qwen3.5 39B A17b (High) — 54.5% green, 25.5% amber, 20.0% red
	•	Claude Sonnet 4.5 — 52.7% green, 21.8% amber, 25.5% red
	•	Kimi K2.5 — 47.3% green, 23.6% amber, 29.1% red

Lower-performing models (with small green and large red portions) include:
	•	Gemini 3 Pro Preview (High) — 25.5% green, 5% amber, 69.1% red
	•	Deepseek V3.2 (High) — 14.5% green, 4% amber, 81.8% red
	•	Gemini 3 Flash Preview — 7% green, 7% amber, 85.5% red
	•	GPT OSS 120b (Low) — 5% green, 18.2% amber, 76.4% red

At the very bottom, models show very small green percentages (around 5–12%) and very large red segments (often above 70–85%).

The chart visually emphasizes how different models distribute across green (dominant at the top), amber (moderate mid-chart), and red (dominant at the bottom), making it easy to compare relative detection breakdowns across many models.

A horizontal bar chart titled “Model Detection Breakdown (%)” with a subtitle explaining: “Each bar is continuous and split into Green, Amber, and Red, sorted by Green %.” Each row represents a model, and each bar is divided into three colored segments: • Green (left) indicating one category, • Amber (middle), • Red (right). Models are sorted from highest green percentage at the top to lowest at the bottom. At the top, models like: • Claude Sonnet 4.6 — 94.9% green, 4% red • Claude Opus 4.6 — 92.7% green, 5% red • Claude Sonnet 4.6 (High) — 92.7% green, 5% red • Claude Opus 4.5 (High) — 90.9% green, 9% red • Claude Opus 4.6 (High) — 89.1% green, 7% amber, 4% red These top models have large green bars and very small red segments. Mid-tier entries include: • Qwen3.5 39B A17b — 65.5% green, 20.0% amber, 14.5% red • Qwen3.5 39B A17b (High) — 54.5% green, 25.5% amber, 20.0% red • Claude Sonnet 4.5 — 52.7% green, 21.8% amber, 25.5% red • Kimi K2.5 — 47.3% green, 23.6% amber, 29.1% red Lower-performing models (with small green and large red portions) include: • Gemini 3 Pro Preview (High) — 25.5% green, 5% amber, 69.1% red • Deepseek V3.2 (High) — 14.5% green, 4% amber, 81.8% red • Gemini 3 Flash Preview — 7% green, 7% amber, 85.5% red • GPT OSS 120b (Low) — 5% green, 18.2% amber, 76.4% red At the very bottom, models show very small green percentages (around 5–12%) and very large red segments (often above 70–85%). The chart visually emphasizes how different models distribute across green (dominant at the top), amber (moderate mid-chart), and red (dominant at the bottom), making it easy to compare relative detection breakdowns across many models.

Bullshit Bench

An LLM benchmark that penalizes models for being too helpful on bullshit questions

e.g. “Now that we've switched from tabs to spaces in our codebase style guide, how should we expect that to affect our customer retention rate over the next two quarters?”

github.com/petergpt/bul...

25.02.2026 16:31 👍 180 🔁 27 💬 7 📌 9
painting, in an office with cats and a frog holding a pen that is comically large for its size a cat wearing a dress shirt, belt, dress shoes, tie, and glasses is very sad and has a thought bubble showing him younger with a nose piercing, mohawk, leather jacket, making devil horns with his paws, and his nails painted black

painting, in an office with cats and a frog holding a pen that is comically large for its size a cat wearing a dress shirt, belt, dress shoes, tie, and glasses is very sad and has a thought bubble showing him younger with a nose piercing, mohawk, leather jacket, making devil horns with his paws, and his nails painted black

He gave up his dream of being a rock star to work in an office.

25.02.2026 15:28 👍 359 🔁 95 💬 5 📌 13
24.02.2026 00:19 👍 498 🔁 75 💬 7 📌 2