Douglas Abreu

vfx cgi ui ux llm ai code

[Google @Google]

Can you hear that? Our Gemini Audio family is getting louder 🔊

We're introducing two of our most expressive audio generation models yet from @GoogleDeepMind: Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS.

[Logan Kilpatrick @OfficialLoganK]

Introducing Gemini 3.8 Flash and Flash-Lite TTS, our new SOTA text to speech model with:

- a new voice design experience
- 2,000+ production ready voices
- voice replication
- support for 100 languages
- voice remixing (soon)
- #1 spot on Hume AI's voice benchmarks

and more!!

[Douglas Abreu @DouglasTAbreu]

@CharaspowerAI vamos pensar como acontece com a gente eu acho que isso é verdade sim

[AndrewOnXYZ @AndrewOnXYZ]

What the actual fuck. Opus 5.5 ultra created this masterpiece. 4 agents and an hour and a half later. Everything from scratch, no AI voice API used...

this is it...

[Ahmet Doğan @Ahmetdoogan]

I tested arrow 2.0

result: PERFECT😱

for me, that represents a 95% success rate in SVG generation

congrats @QuiverAI and thanks for adding it @paper

[Qwen @Alibaba_Qwen]

⚡ Meet Qwen-Audio-3.1! ASR, TTS & Realtime are fully upgraded, joined by two new models: TTS-Next for audio creation and ASR-Next for audio understanding.

Five models, one complete audio stack: understanding, generation, interaction & creation.

Plus big price cuts across the lineup: TTS ~70% off, Realtime ~85% off, and ASR up to 95% off.

Highlights: 🥳
- ASR: stronger multilingual & dialect recognition, plus native polishing that auto-removes fillers & repetitions for cleaner, more logical transcripts.
- ASR-Next: supports multi-speaker ASR with speaker labels, timestamps & aligned transcripts, and understands emotions, ambient & machine sounds for sound captioning, event localization, audio QA & reasoning.
- TTS: multilingual & dialect synthesis with natural cross-lingual voice transfer; control emotion, speed & style via simple instructions.
- TTS-Next: unified LM + diffusion framework generating voice, sound effects & background audio in one pass for audiobooks, podcasts, games & ads.
- Realtime: speak & listen at once with anytime interruption, just like a real call; it even slows down and responds empathetically when it senses a low mood.

Unlock the full potential of Qwen-Audio-3.1! 👇
- Blog: https://t.co/e0M8wvj8Kg
- Qwen-Audio-3.1-ASR:
https://t.co/88lPQTPxyz
- Qwen-Audio-3.1-Realtime:
https://t.co/Y63sdtK2AC
- More APIs: coming soon @qwen_cloud

[SciTech Girl @scitechgirl]

🚨 SCIENTISTS FOUND LIGHT INSIDE HUMAN CELLS

Researchers discovered that tiny parts inside our cells may communicate using microscopic flashes of light called biophotons. These signals are invisible to the human eye, but scientists believe they could help cells coordinate, repair damage, and respond to stress.

The discovery is changing how experts understand the human body — suggesting that deep inside us, cells may be “talking” through light itself.

Source:
Kučera, O., et al. *Biophotons and mitochondrial communication in living cells*. Photochemical & Photobiological Sciences.

[Markus J. Buehler @ProfBuehlerMIT]

Fascinating AI swarm dynamics: a few agents spontaneously emerge as highly connected hubs, while most remain locally connected. The swarm develops a strongly heterogeneous interaction topology with a long-tailed degree distribution - an emergent organizational structure arising from initially decentralized local interactions. There is no central planner assigning roles; the swarm builds its own coordination architecture, with information brokers and increasingly global integration emerging from local behavior.

[Dev Ed @developedbyed]

Opus 5.5 tries to capture the feeling of childhood wonder in a one-minute animated short

Took about 1 hour and 20 minutes, this is the one shot. Some parts could definitely be smoothed out but wow what a great starting point.

[Jimmy Apples 🍎/acc @apples_jimmy]

Asked opus 5.5 to make a video of openai solving Navier-Stokes.

Very nice model, I still prefer Astra and 5.6 as my main drivers as I do a lot of research heavy work but 5.5 Opus is the first time since 4.7 that I want to heavily run Claude.

It has great taste.

[Paweł Huryn @PawelHuryn]

I finally tested GPT-6 Sol on a real work. 2 repos. 105 hidden bugs. Find and fixed what you can.

It looks like a huge degradation. The results:
- GPT-6 Astra (max): 45
- GPT-5.6 Sol (max): 43.5
- Opus 5.5 (max): 41.7
- Muse Spark 1.3 (max): 32.2
- GPT-6 Sol (max): 29.3

Until you measure the cost (API-equivalent):
- GPT-6 Astra (max): $33.04
- GPT-5.6 Sol (max): $95.25
- Opus 5.5 (max): $58.53
- Muse Spark 1.3 (max): $18.11
- GPT-6 Sol (max): $9.93

More effort levels (xhigh, high, medium, low) dropping in this thread today 🧵

[OpenAI Developers @OpenAIDevs]

We’ve improved prompt caching in the API for GPT-6, helping agents run faster and cost less.

Higher cache-hit rates by default mean more input tokens benefit from cached-input discounts of up to 90%.

[Douglas Abreu @DouglasTAbreu]

@thsottiaux o google está com um boca enorme e vai devorar esses fofinhos

[DreW @devteamdrew]

Made with @claudeai Opus 5.5

[Kevin Ngo @kevin_t_ngo]

Claude Opus 5.5 drew every frame of this animation in JavaScript.

Everyone in town sends Claude their requests, but one girl sends a question instead: "What do you love?"

[Carlos Santana @DotCSV]

🔴 ¡OPENAI ANUNCIA GPT-6 SOL y LUNA!

La gama media (Sol) y baja (Luna) de modelos -tras la salida de los modelos Terra- evolucionan a GPT 6!

Además, bajada de precios en la API

[Douglas Abreu @DouglasTAbreu]

@thsottiaux o que as pessoas querem é reset não modelos mais inteligentes, voce nunca será esquecido por reinventar o RESET

[Chetaslua @chetaslua]

Claude opus 5.5 made this launch video in JS

gave the original tweet and prompted it to code it no external music , sound ,image assets everything is code

https://t.co/rEZNcMsrBw

[Carlos Santana @DotCSV]

Sé que es un chascarillo, pero también sé que mucha gente lo va a malinterpretar, así que

1) no es un modelo frontera
2) el red teaming de este modelo (el probarlo) es anterior a cualquier cosa que haya sucedido en las últimas 2 semanas
3) ambos dijeron que frenar no significaba parar de sacar modelos, sino darse más tiempo para evaluar a sus modelos más potentes

[Brais Moure @MoureDev]

12 de septiembre: Hay que frenar a la IA!
22 de septiembre: Chavales, nuevo modelo, pero tranquis, que está bien probado!

Han frenado 10 días.