The 1X Neo humanoid robot aims to bring “physical AI” into homes as a household helper. Capable of tasks like dishwashing and folding laundry, Neo currently relies on remote human operators using VR. 1X plans for full autonomy by 2026, gathering real-world data from users to train its AI—raising privacy and safety concerns.

Eli Lilly is partnering with Nvidia to build a supercomputer powered by over 1,000 Blackwell GPUs, set to go online by January using 100% renewable energy. The collaboration aims to accelerate drug discovery and development, harnessing AI’s full potential to revolutionize how new medicines are created and brought to market.

While AI accelerates scientific breakthroughs like protein engineering and longevity research, it may also threaten human well-being. Rising deaths among younger adults—linked to joblessness and social isolation—suggest AI-driven unemployment could deepen despair and reduce life expectancy, even as older generations benefit from AI-enhanced healthcare and longevity advances.

OpenAI has developed GPT-4b micro, its first biology-focused AI, capable of designing proteins that can turn regular cells into stem cells. In collaboration with Retro Biosciences, the model improved Yamanaka factors up to 50-fold. The project signals OpenAI’s move into scientific discovery and its potential role in achieving artificial general intelligence.

Humanity is rapidly approaching digital superintelligence. AI is accelerating scientific progress and productivity, promising immense societal change and abundance. The key challenge ahead is ensuring safety, alignment, and fair access as intelligence becomes ubiquitous and transformative.

AI is already transforming major industries from healthcare and education to finance and agriculture. This article describes how AI improves efficiency and outcomes, outlines key benefits and required skills, and addresses challenges and concerns unique to each field.

Read about FastNet, the next-generation AI weather prediction model developed by the Alan Turing Institute and the UK Met Office. FastNet, faster, more precise and requiring lower energy than physics-only models, will combine machine learning with physics-based models, to revolutionize weather prediction, renewable energy management and climate science.

AI is gaining pace in healthcare with FDA determined to use it in order to speed up the final stages of making a drug or medical device but also to review chemicals and other ‘concerning ingredients’ in food. This is a major breakthrough as long as the model is the right one and is trained by correct data.

Are AI systems getting too smart for us to measure? Some of them are already getting high scores on many PhD-level challenges. A new evaluation called ‘Humanity’s Last Exam’ has been designed to determine whether AI systems have become so good as to be considered ‘world-class’ oracles. ‘Humanity’s Last Exam’ is as promising as intimidating.

Across the spectrum of uses for artificial intelligence, one stands out: scientific discovery. As AI is fed with enormous amount of scientific data, its most significant opportunity on the horizon lies in accelerating scientific development to generate new drugs to combat disease, new agriculture to feed the world’s population and new materials to unlock green energy. This is where the startup Lila is targeting.

Artificial Intelligence, through the use of intelligent coding agents, is changing the coding landscape by automating routine tasks and empowering developers to work more creatively and efficiently. While it is here augment software developers’ abilities and not to replace them, it fundamentally transforms how coding is done, introducing a new "vibe" to the process. (subscription required)

A breakthrough AI diagnostic tool is entering the medical field! This tool is significantly surpassing human doctors in diagnostic ability and is set to reduce medical errors and expand access to expert care worldwide. This will could transform diagnostics by improving accuracy and efficiency.

Can AI ‘deskill’ people if used too much? This is what a study published in ‘The Lancet Gastroenterology & Hepatology’ claims about doctors who lose their ability to detect adenomas in colonoscopies following long-term use and thus dependency on AI. This outcome underlines the importance of a balanced approach towards AI usage in medicine, and other areas of human activity.

At MIT Energy Initiative’s Spring Symposium titled "AI and energy: Peril and promise" in May 2025, researchers assesed the dual role of AI as both a problem, due to its soaring energy consumption, and a potential solution, via the proposal of optimization solutions for energy production and consumption. The article sheds light to this energy conundrum but also to the possible solutions and challenges in the path solving the problem.