Cosine just showed off Genie, a new fully autonomous AI software engineer that broke the high score for SWE-Bench, by over 10%

When it makes a mistake, it iterates, re-plans, and re-executes until it fixes the problem. Wild.

Nima Owji spotted an early look at Grok 2.0, which is coming to beta mode "soon" according to Elon.

It's reportedly better at code, writing, generating new, and can generate images with the new open source FLUX.1 model.

OpenAI and the authors of SWE-bench collaborated to redesign the popular software engineering benchmark and release ‘SWE-bench Verified’

It addresses issues in the original benchmark, like overly specific unit tests and unreliable development environments

Forget drones, this street-smart robot could be future of local deliveries

https://www.foxnews.com/tech/forget-drones-street-smart-robot-could-future-local-deliveries

OpenAI collaborated with The Met museum to create an interactive installation, building a custom ChatGPT model and using Sora to bring a 1930s New York socialite to life

Checked out an AI News Presenter demo on Reddit using Flux Dev for character creation, RunwayML for animation, and Synthesia IO for voice. The integration is seamless, showcasing the growing role of AI in media.

Astribot S1 just released a new demo—watch how it handles everyday tasks with incredible agility and precision.

Here's Gen-3 Alpha Image to Video invading a house.

This AI-generated video showing a variety of expressions. 🎥
The detail captured in each frame is impressive.

AGIBOT, a China-based robotics startup, unveiled a family of 5 advanced humanoid robots.

The robots are specifically designed for diverse tasks — from household chores to industrial operations, directly challenging Elon and Tesla’s Optimus

Unitree, another high-performance robot manufacturer from China, showcased its new G1 humanoid robot.

The updated bot is nearing "mass production" and comes at a price tag of only $16,000

This thing has insane whole-body control

Luma Labs just released Dream Machine 1.5, a major upgrade to their current AI video generation model.

The update boasts higher-quality text-to-video, smarter prompt understanding, and better image-to-video capabilities.

A Swiss startup called FinalSpark created AI “biocomputers“ made from real human brain cells to create more energy-efficient AI.

The craziest part is that they're rentable for just $500 a month.

Midjourney released a new unified web-based AI image editor.

It combines several of the advanced image generation features to seamlessly modify and extend generated images

Neuralink released its progress report on its Second human patient, Alex.

TLDR:

- Discharged the next day
- Broke the prev. world record for BCI cursor control on day one
- Can play Counter-strike 2 using just the Neuralink + his brain (no controller) 🤯

CodeSignal launched Conversation Practice, an AI-powered tool for simulating workplace convos and providing personalized feedback

This is like GPT-4o voice for practicing job interviews, leadership, and communication skills

ETH Zurich and Disney developed an AI system for enhancing humanoid robot movement (and giving it the ability to dance!)

The system bridges the gap between animation and robotics, helping humanoids move more naturally and adapt better to new situations