Meta is releasing a standalone mobile app for its ChatGPT competitor, Meta AI.

There's a Discover feed that shows interactions that others (including your IG/FB friends) are having with the assistant. Meta tells the idea is to demystify AI and show “people what they can do with it."

OpenAI is working on a similar feed for ChatGPT.

Xiaomi MiMo-7B: a 7B reasoning model series trained from scratch, outperforms 32B+ baselines on math and code via dense RL

- pretrained on 25T tokens w/ multi-token prediction
- RL rewards from rule-verifiable math/code tasks
- cold-start RL model (MiMo-7B-RL-Zero) hits 93.6% MATH-500, 49.1% LCB v5
- SFT→RL variant matches OpenAI o1-mini
- also open: base + SFT checkpoints
- seamless rollout engine: 2.29× faster RL training
- vLLM + MTP inference ready
- strong AIME 2025 (55.4%) and LCB v6 (49.3%) results

Microsoft is getting ready to host Elon Musk’s Grok AI model. Microsoft has been in discussions with xAI to make Grok AI available on Azure's AI Foundry service.

In recent weeks Microsoft has been in discussions with xAI to host the Grok AI model and make it available to customers and Microsoft’s own product teams through the Azure cloud service.

The move could prove controversial internally and further inflame tensions with Microsoft’s partner OpenAI.

Microsoft Introduced Phi-4-reasoning, adding reasoning models to the Phi family of SLMs.

The model is trained with both supervised finetuning (using a carefully curated dataset of reasoning demonstration) and Reinforcement Learning.

- Competitive results on reasoning benchmarks with much larger top-tier models up to DeepSeek R1.

- Strong performance on new tests released after data collection (AIME 2025, HMMT).

- Reasoning transfers/generalizes well to new domains even with only SFT (e.g. k-SAT, Mae Solving, Calendar Planning, etc.)

- Retains and often significantly improves general-purpose capabilities (e.g. instruction following).

HuggingFace Phi-4-reasoning
HF Phi-4-reasoning-plus
Hf Phi-4-mini-reasoning

Carnegie Mellon University started company with only AI employees

They got OpenAI, Gemini, Anthropic, etc models and gave them job roles. They were retarded and costed a fuckton. Claude 3.5 was the best employee and only did 24% of its tasks.

Paper.

Google released an updated Gemini 2.5 Pro

With this update, you can create even more complex web apps from a single prompt.

Immersive Translate is the ultimate tool for seamless bilingual translations across the web, videos, PDFs, and EPUBs. With support for over ten top translation engines like DeepL, OpenAI (ChatGPT), and Google Translate, it brings effortless language learning and content accessibility. From translating foreign language web pages to real-time bilingual subtitles for videos and even handling complex PDF documents, Immersive Translate is a game changer for multilingual users. Use the special code blockcw Join here https://dada.link/CbUdUd

Watch full review here
https://youtu.be/TUfX8TvHKgM

OpenAI launched a new initiative "OpenAI for Countries" to help governments build democratic AI infrastructure through partnerships that include building in-country data centers, providing customized ChatGPT to citizens, evolving security and safety controls, and raising national startup funds, with a goal to pursue 10 projects with individual countries or regions in the first phase as part of their Stargate project.

Meta introduced LlamaFirewall: An open source guardrail system for building secure AI agents

Mitigates risks such as prompt injection, agent misalignment, and insecure code risks through three powerful guardrails.

GitHub.
Paper.

Meta Locate 3D: A model for accurately localizing objects in 3D environments.

Learn how Meta Locate 3D can help robots accurately understand their environments and interact more naturally with humans.

You can download the model and dataset at https://go.fb.me/je2t32

OpenAI just released HealthBench — a new eval for AI systems for health.

Developed with 262 physicians who have practiced in 60 countries.

Create applications IN SECONDS — a new neural network Appalchemy has been released.

AI ITSELF will come up with the design and features of the application based on your description — all you have to do is admire the finished layout or ask for edits.

It can be used for free http://appalchemy.ai/

Meta just released new models, benchmarks, and datasets that will transform the way researchers approach molecular property prediction, language processing, and neuroscience.

1. Open Molecules 2025 (OMol25): A dataset for molecular discovery with simulations of large atomic systems.

2. Universal Model for Atoms: A machine learning interatomic potential for modeling atom interactions across a wide range of materials and molecules.

3. Adjoint Sampling: A scalable algorithm for training generative models based on scalar rewards.

4. FAIR and the Rothschild Foundation Hospital partnered on a large-scale study that reveals striking parallels between language development in humans and LLMs.

Nvidia has made the world happy: the most powerful speech recognition model is absolutely FREE and right in your browser!

Forget about paid services! This new product will decipher anything without further ado: videos, podcasts, lectures, chatter with friends and even the cacophony of the streets. It understands sarcasm (well, almost), places commas like a hereditary philologist, and marks time periods like an experienced director.

It doesn't matter what the input is: song lyrics, a stream of numbers or a multi-hour discussion - this beast weighs only 2 GB and is ready to work.

Test it online in a browser or download it to your computer

https://huggingface.co/spaces/nvidia/parakeet-tdt-0.6b-v2

OpenAI introduced AI agent codex.

it is a software engineering agent that runs in the cloud and does tasks for you, like writing a new feature of fixing a bug.

U can run many tasks in parallel. Starting to roll out today to ChatGPT pro, enterprise, and team users.

AI now makes 4K video: in the SkyReels V2 update, you can make videos in the highest quality

Voice, script, lip sync, music and video - all in one service and without restrictions.

https://www.skyreels.ai/home

RECURSE is the world's first track born from a collaboration between a human and quantum AI.

British artist ILĀ and startup Moth used the Archaeo machine learning platform, which runs on a quantum computer. Unlike common AI generators such as Suno and Udio, which are trained on huge amounts of data, Archaeo trained only on ILĀ's music. This allowed quantum algorithms to capture the unique nuances of the artist's work and create more "human" music, avoiding issues with ethics and copyright.

Adobe introduced HUMOTO, 4D dataset for human-object interaction, developed with a combination of wearable motion capture, SOTA 6D pose estimation vision models, LLM, and the professional refining works of multiple animation studios.

HUMOTO features:
1. Over 700 diverse daily activities
2. Interactions with 60+ objects, 70+ articulated parts.
3. Fine-grained text annotations
4. Detailed hand and finger movements.

Google is launching their own coding agent, Jules, at I/O today

It lets you make changes to your GitHub repos with English prompts in a VM using Gemini 2.5 Pro. It's Google’s version of Devin.

Here's the leaked official repo of prompts it can do.

What if I told you this viral music video was made with zero filming, zero crew… and just one AI tool? See full tutorial!

‼️ HOLLYWOOD IS IN IT! Google's new Veo 3 neural network can now not only create cinematic videos, but also voice them with real people's voices

▸ The recently updated Veo 3 can now generate sound effects, background noise, and even dialogue based on your requests

▸ The model also perfectly perceives complex instructions and accurately reproduces the sequence of events, simplifying the creation of realistic scenes with reliable movement and sound

Considering the rapid pace of development of neural networks, it may be possible to make films in the future at all... 🤔