Tootfinder

Sources: OpenAI begins offering its chatbot ads to dozens of advertisers, initially charging per ad view, not per ad click, and asks them for <$1M commitments (The Information)
https://www.theinformation.com/articles/openai-lines-adver…

OpenAI Lines Up Advertisers, Reveals Key Details Ahead of Ads Launch
OpenAI has started offering its new chatbot ads to dozens of advertisers, according to two people familiar with the matter. To start off, the company is charging on the basis of ad views as opposed to per-ad clicks that are typical at large sellers of search ads like Google and Amazon, the ...

Easy Adaptation: An Efficient Task-Specific Knowledge Injection Method for Large Models in Resource-Constrained Environments
Dong Chen, Zhengqing Hu, Shixing Zhao, Yibo Guo
https://arxiv.org/abs/2512.17771 https://arxiv.org/pdf/2512.17771 https://arxiv.org/html/2512.17771
arXiv:2512.17771v1 Announce Type: new
Abstract: While the enormous parameter scale endows Large Models (LMs) with unparalleled performance, it also limits their adaptability across specific tasks. Parameter-Efficient Fine-Tuning (PEFT) has emerged as a critical approach for effectively adapting LMs to a diverse range of downstream tasks. However, existing PEFT methods face two primary challenges: (1) High resource cost. Although PEFT methods significantly reduce resource demands compared to full fine-tuning, it still requires substantial time and memory, making it impractical in resource-constrained environments. (2) Parameter dependency. PEFT methods heavily rely on updating a subset of parameters associated with LMs to incorporate task-specific knowledge. Yet, due to increasing competition in the LMs landscape, many companies have adopted closed-source policies for their leading models, offering access only via Application Programming Interface (APIs). Whereas, the expense is often cost-prohibitive and difficult to sustain, as the fine-tuning process of LMs is extremely slow. Even if small models perform far worse than LMs in general, they can achieve superior results on particular distributions while requiring only minimal resources. Motivated by this insight, we propose Easy Adaptation (EA), which designs Specific Small Models (SSMs) to complement the underfitted data distribution for LMs. Extensive experiments show that EA matches the performance of PEFT on diverse tasks without accessing LM parameters, and requires only minimal resources.
toXiv_bot_toot

YouTube CEO Neal Mohan says YouTube TV plans to launch "fully customizable" multiview for non-sports channels and channel packages focused on various topics (Emma Roth/The Verge)
https://www.theverge.com/news/864566/youtube-tv-custom-multiview-chan…

YouTube TV will launch custom multiview and channel packages ‘soon’
YouTube TV is getting ready to launch customizable multiview, along with genre-specific channel packages. Both of these features are coming “soon,” according to YouTube CEO Neal Mohan.

Layoffs hit Vimeo for the second time since September 2025, when it laid off 10% of its workforce; Bending Spoons acquired Vimeo for around $1.38B in November (Business Insider)
https://www.businessinsider.com/vimeo-laying-off-staff-after-bi…

Vimeo is laying off staff globally after its $1.38 billion sale to Bending Spoons
Vimeo is letting go of workers globally in a new round of layoffs after selling its business to tech-holding company Bending Spoons.

NYT launches Crossplay, its first two-player competitive word game, as a standalone iOS and Android app, as it seeks a bigger share of the mobile gaming market (Alex Weprin/The Hollywood Reporter)
https://www.hollywoodreporter.com/business

The New York Times Is Launching Its First Two-Player Word Game: Crossplay (Exclusive)
The game will be familiar to Scrabble or Words with Friends players, but the Times is adding a few twists of its own.

After backlash, Microsoft deleted a blog post written in November 2024 that critics said encouraged developers to pirate Harry Potter books to train AI models (Ashley Belanger/Ars Technica)
https://arstechnica.com/tech-policy/20

Microsoft deletes blog telling users to train AI on pirated Harry Potter books
The now-deleted Harry Potter data set was "mistakenly" marked public domain.

Cate Blanchett, Cyndi Lauper, and 800 creatives launch "Stealing Isn't Innovation" to protest unauthorized AI training, backed by RIAA, SAG-AFTRA, and others (Mia Sato/The Verge)
https://www.theverge.com/ai-artificial-int…

Hundreds of creatives warn against an AI slop future
The Stealing Isn’t Innovation campaign — supported by artists like Cate Blanchett, Cyndi Lauper, and George Saunders — warns of a future filled with AI slop.

Sources: banks including Wells Fargo, BNP Paribas, and HSBC began selling an extra $8.2B of short-term loans to help Netflix fund its WBD purchase (Bloomberg)
https://www.bloomberg.com/news/articles/2026-01-20/banks-selling-8…

Atlas Obscura plans to launch Obscura Society, a WebXR-based lounge offering users 3D avatars for voice chats, portals to other VR apps, and more, in early 2026 (Janko Roettgers/The Verge)
https://www.theverge.com/column/847182/atlas-obscura-society-3d-virtual-…

How Atlas Obscura plans to embrace the 3D web
The Obscura Society will be a virtual lounge that Atlas Obscura community members can visit to hang out with the help of 3D avatars, have voice chats, and explore places from the site’s massive catalog together.

Google returns as a partner service for Movies Anywhere, after leaving the service during a feud with Disney in October and November (Will Sattelberg/9to5Google)
https://9to5google.com/2025/12/18/google-youtube-return-movies-anywhere-how-to-re-…

Google and YouTube return to Movies Anywhere, here’s how to re-link your account
As promised, both Google Play and YouTube have returned to Movies Anywhere, here's how to restore your account link.