✔️ Interactive visualizations have been added to ChatGPT, making it easy and visually clear to study mathematics and physics.
OpenAI has introduced a new feature in ChatGPT — dynamic graphical explanations for over 70 school and scientific concepts. Users can now change variables during a conversation and immediately see how these changes affect graphs and formulas. This approach makes learning more interactive and understandable.
Currently, this feature is primarily designed for school curricula and university courses. Available topics include binomial squares, exponential decay, Ohm’s law, compound interest, and trigonometric identities.
This functionality is now accessible to all users worldwide. In the future, OpenAI plans to expand the list of educational modules and add new disciplines.
openai.com
—
Nvidia will allocate an entire gigawatt of computing power to the startup Miry Murtadi for AI development.
Thinking Machines Lab — a project founded by former OpenAI leader Miry Murtadi — has signed a strategic agreement with Nvidia. Under the terms, the startup will gain access to at least one gigawatt of capacity based on the latest Vera Rubin systems. These resources are intended for training proprietary AI models. Cluster deployment is scheduled for early next year.
Additionally, Nvidia increased its stake in the project: it previously participated in a $2 billion funding round, valuing the company at $12 billion.
Together, the companies will develop infrastructure solutions for training and deploying neural networks to provide businesses and the scientific community with cutting-edge AI solutions.
blogs.nvidia.com
—
Amazon has introduced a new rule: all code generated with AI must now be reviewed by experienced engineers before publication.
This decision was made following a series of serious incidents characterized by the company as failures with significant consequences. Amazon management aims to increase control over automated software generation: senior developers now need to approve each change.
At a recent meeting, top executives acknowledged that the industry has yet to develop reliable safeguards when working with code generators. For now, manual expert review remains the primary method to prevent large-scale system failures and production outages.
ft.com
—
Nvidia is working on launching NemoClaw — an open platform for creating and deploying autonomous AI agents without dependence on specific Nvidia hardware.
The project is planned to be open-source, allowing developers to easily implement their solutions outside the CUDA ecosystem. An official announcement is expected at the upcoming conference in San Jose. Early access negotiations are underway with major companies such as Salesforce, Google, Cisco, Adobe, and CrowdStrike.
For Nvidia, releasing this platform is a significant step forward. It helps maintain leadership in AI infrastructure amid active development of custom chips by major industry players.
wired.com
—
Cortical Labs announced the launch of its first biological data centers: the first centralized cluster opened in Melbourne, with plans for a second in Singapore. Instead of traditional servers, they use unique biocomputers called CL1.
These systems are based on neurons — grown from human cells and placed on specialized chips. They interact with hardware via electrical signals, and software interprets the cells’ reactions as computational results.
While not yet fully competing with classical processors, progress is impressive: around 120 such bio-modules are already operational in Melbourne. In Singapore, plans are to increase their number to a thousand as infrastructure expands.
bloomberg.com
Created with n8n:
https://cutt.ly/n8n
Created with syllaby:
https://cutt.ly/syllaby