AI Valley
Posts
OpenAI’s Sora leaked by protestors

OpenAI’s Sora leaked by protestors

PLUS: Google is connecting Spotify to Gemini AI

Barsee
27 Nov

_{Sign up}_|_{Follow us on X}_|_Sponsor

Together with

Howdy! It’s Barsee again.

Happy Wednesday, AI family, and welcome back to AI Valley.

In today’s edition:

OpenAI’s Sora leaked by protestors
Google is connecting Spotify to Gemini AI
Plus trending AI tools, posts, and resources

Ready, set, go…

SORA

OpenAI’s Sora leaked by protestors

A group of "early testers, red teamers, and creative partners" has leaked OpenAI's Sora AI video model via Hugging Face to protest what they say is unpaid development work. The group says OpenAI asked hundreds of artists to test the system, find errors, and provide feedback without fair compensation.

Here's what you need to know:

The leaked version generates 10-second, 1080p videos and looks like a faster, "turbo" variant. The code also suggests style controls and restricted customization options.
Many Sora generated videos surfaced on X following the leaks. However, the leaked frontend stopped working after few hours, likely due to OpenAI revoking access.
The leak was part of a protest by artists who claim OpenAI exploited them for unpaid testing and feedback, providing only "artistic credibility" in return.
They also criticized OpenAI for tightly controlling early access to Sora, requiring content sharing approval, favoring select creators, and misrepresenting the model's true capabilities.

Why it matters:

The Sora leak highlights growing tensions between artists and AI developers over fairness, transparency, and the ethics of creative contributions in AI projects.

TOGETHER WITH SURFSHARK

V1: Out-of-this-world Black Friday VPN deal

This Black Friday, level up your online security with an unbeatable deal from Surfshark! For just $1.99/month, you’ll get 2 years of top-tier VPN protection and 4 extra months absolutely FREE.

Safeguard your privacy, stream securely, block annoying ads, and stay protected on public Wi-Fi—all on unlimited devices. Don’t miss out on this limited-time offer.

SIDE UPDATES

🤗 Hugging Face unveils compact yet powerful Vision Language Model

Hugging Face has launched SmolVLM, a family of small vision-language models designed for accessibility without sacrificing performance. Available in three variants: SmolVLM-Base, SmolVLM-Synthetic, and SmolVLM-Instruct, this model requires just 5.02 GB of GPU RAM for inference and handles longer sequences with a 16k token context window. It outperforms larger models like Qwen2-VL, delivering 3.3-4.5x faster prefill throughput and 7.5-16x faster generation throughput across various vision-language tasks.

✍️ Anthropic introduces Custom Writing Styles for Claude AI

Anthropic has launched a new feature for Claude AI, allowing users to customize the chatbot’s responses based on their preferred writing style. Users can choose from three preset styles: Formal, Concise, and Explanatory or create a custom style by providing sample content and guidance.

💰️ SoftBank to Invest $1.5bn in OpenAI Stake Boost

SoftBank is set to boost its stake in OpenAI with a $1.5 billion stock buyback plan, following its $6.6 billion investment, bringing OpenAI’s valuation to $157 billion. Led by CEO Masayoshi Son, this move allows SoftBank to acquire a larger share, while also offering OpenAI employees an opportunity to sell stock.

🎬 IMAX leverages AI for Global Content Reach with over 140+ languages

IMAX has teamed up with Dubai-based AI startup Camb.AI to introduce real-time language translation in theaters globally. The collaboration focuses on localizing IMAX’s original content and documentaries into 140 languages with high-quality translations that maintain the emotional depth of the original performances, offering a more efficient alternative to traditional dubbing.

GEMINI

Google is connecting Spotify to Gemini AI

Google is rolling out a new Spotify extension for its Gemini assistant, allowing users to control their music playback using natural language commands.

Here's what you need to know:

Users can now directly control Spotify through Gemini, browsing and playing music with voice or text commands by referencing songs, artists, albums, playlists, or activities.
The extension currently only supports basic playback controls and does not allow for creating playlists or radio stations.
To activate the feature, users must link their Spotify account to their Google account and enable Gemini Apps Activity, which stores AI queries for up to 72 hours.
The feature is in early stages and is limited to English in the main Gemini app, with no availability on Google Messages, Gemini web, or iOS devices.

Why it matters:

This integration brings more convenience to Spotify users, allowing hands-free control via Gemini, but its limited features and availability hint at future updates that could offer more advanced music management features.

TRENDING TOOLS

Pickle AI > Your digital twin for seamless video calls [Link]
Screenshot-to-Code > A simple tool to convert screenshots, mockups and Figma designs into clean, functional code using AI [Link]
Spiral > Automate repetitive creative work, in your voice [Link]
Axon Data > Connects multiple data sources to find actionable insights and plan for your business [Link]

THINK PIECES / RESOURCES

Getting started with AI: Good enough prompting [Link]
How you can make AI your full stack developer [Link]
This simple AI hack makes ChatGPT 3x smarter [Link]
A tiny robot kidnapped 12 bigger robots from a Shanghai showroom. [Link]
This course teaches you how to set up and use Ollama to build powerful AI applications locally [Link]