digestdaily-digestai-newstrending
🌅 AI Daily Digest — April 28, 2026
Today: 8 new articles, 5 trending models, 5 research papers
This article was generated by Daily Neural Digest's autonomous neural pipeline — multi-source verified, fact-checked, and quality-scored. Learn how it works
Data Pulse
- 8 news articles
- 5 trending models
- 5 research papers
- Cheapest GPU: RTX 4000Ada at $0.02/hr
- 3 new AI jobs
Today's News
- 4TB of voice samples just stolen from 40k AI contractors at Mercor Mercor.io Corporation, a San Francisco-based AI hiring startup, suffered a data breach compromising 4TB of voice samples from approximately 40,000 AI trainers employed as contractors. The breach highlights vulnerabilities in AI contractor data security.
- China blocks Meta's acquisition of AI startup Manus Chinese regulators have blocked Meta's acquisition of Manus, a Beijing-based AI startup, halting Meta's expansion in generative AI agent development. The decision marks a significant regulatory setback for Meta's strategic goals.
- Elon Musk and Sam Altman are going to court over OpenAI’s future Elon Musk and Sam Altman face a legal showdown in Northern California, escalating a years-long dispute over OpenAI’s governance and future direction. The case could reshape the company’s trajectory.
- Google employees ask Sundar Pichai to say no to classified military AI use Over 600 Google employees signed a letter urging CEO Sundar Pichai to prohibit the use of Google’s AI models for classified military applications. The demand reflects growing internal concerns about ethical AI deployment.
- Microsoft Presents 'TRELLIS.2': An Open-Source, 4b-Parameter, Image-To-3D Model Microsoft unveiled TRELLIS.2, an open-source image-to-3D model with 4b parameters, offering improved efficiency and fidelity in asset generation. The model leverages 16× spatial compression for enhanced performance.
- OpenAI available at FedRAMP Moderate OpenAI has achieved FedRAMP Moderate authorization for its ChatGPT Enterprise platform and core API, enabling broader U.S. government adoption of its generative AI models. This certification expands its enterprise accessibility.
- OpenAI’s New GPT-5.5 Powers Codex on NVIDIA Infrastructure OpenAI deployed GPT-5.5 to power Codex, its AI coding tool, marking a significant upgrade from earlier GPT versions. NVIDIA is already utilizing the model for its infrastructure.
- The Download: DeepSeek’s latest AI breakthrough, and the race to build world models DeepSeek, a Chinese AI firm, previewed its V4 large language model, advancing the global competition to develop world models. The release underscores rapid progress in large-scale AI research.
Trending Models
| Model | Task | Likes |
|---|---|---|
| meta-llama/Llama-3.1-8B-Instruct | text-generation | 5763 |
| openai/gpt-oss-20b | text-generation | 4567 |
| deepseek-ai/DeepSeek-R1 | text-generation | 13313 |
| Qwen/Qwen3-0.6B | text-generation | 1212 |
| Qwen/Qwen2.5-7B-Instruct | text-generation | 1243 |
Research
- Spend Less, Fit Better: Budget-Efficient Scaling Law Fitting via Active Experime — Sijie Li, Shanda Li, Haowei Lin. Scaling laws are used to plan multi-million-dollar training runs, but fitting those laws can itself cost millions.
- Relaxation-Informed Training of Neural Network Surrogate Models — Calvin Tsay. ReLU neural networks trained as surrogate models can be embedded exactly in mixed-integer linear programs (MILPs), enabling global optimization over the learned function.
- Neural Recovery of Historical Lexical Structure in Bantu Languages from Modern D — Hillary Mutisya, John Mugane. We investigate whether neural models trained exclusively on modern morphological data can recover cross-lingual lexical structure consistent with historical reconstruction.
- Zero-Shot Morphological Discovery in Low-Resource Bantu Languages via Cross-Ling — Hillary Mutisya, John Mugane. We present a method for discovering morphological features in low-resource Bantu languages by combining cross-lingual transfer learning with unsupervised clustering.
- Aligning Dense Retrievers with LLM Utility via DistillationAligning Dense Retrie — Rajinder Sandhu, Di Mu, Cheng Chang. Dense vector retrieval is the practical backbone of Retrieval- Augmented Generation (RAG), but similarity search can suffer from precision limitations.
GPU Deals
| GPU | Price | Provider |
|---|---|---|
| RTX 4000Ada | $0.02/hr | Vast.ai |
| Tesla V100 | $0.02/hr | Vast.ai |
| RTX 5060 Ti | $0.04/hr | Vast.ai |
View full GPU pricing dashboard
AI Jobs
- Mid Senior AI Video Artist at EverAI (Remote)
- Senior Technical Lead at Calibrate Group AI (Remote)
- ML Engineer at Trunk Tools, Inc. (Remote)
Community Events
New this week:
- Google I/O 2026 (Mountain View, USA)
- Springing into AI: PyTorch Conference Europe and ICLR 2026 (Online)
- ACL 2026 (Online)
- Microsoft Build 2026 (Seattle, USA)
- MLOps Community Weekly Meetup (Online (Zoom))
daily-digestai-newstrendingresearch
Was this article helpful?
Let us know to improve our AI generation.
Previous Article🌅 AI Daily Digest — April 27, 2026Next Article 🌅 AI Daily Digest — April 29, 2026
Related Articles
Digest
🌅 AI Daily Digest — April 29, 2026
Today: 9 new articles, 5 trending models, 5 research papers
4 minabout 3 hours agodaily-digest
Digest
🌅 AI Daily Digest — April 27, 2026
Today: 6 new articles, 5 trending models, 5 research papers
4 min2 days agodaily-digest
Digest
🌅 AI Daily Digest — April 26, 2026
Today: 7 new articles, 5 trending models, 5 research papers
4 min3 days agodaily-digest