India will take another crack at a moon landing on Chandrayaan 3. Indeed, that's the primary focus: The new mission carries a lander and rover but no orbiter.
Generative AI, a technology that can create human-like responses to user prompts, is being tested by the US military for the first time. The Pentagon is running an eight-week exercise with five large-language model (LLM) platforms, such as Scale AI’s Donovan, that are trained on huge amounts of internet data. The LLMs can help...
The arrangement of electrons in matter, known as the electronic structure, plays a crucial role in fundamental but also applied research such as drug design and energy storage. However, the lack of a simulation technique that offers both high fidelity and scalability across different time and length scales has long been a...
Here’s a puzzle: We’ve seen a lot of papers claiming you can use one language model to generate useful training data for another language model. But…by the data processing inequality, we shouldn’t expect to be able to create new information that wasn’t in the first model’s training set.
Mastodon fixes major security flaw that could have allowed system hijacking
'The flaws include CVE-2023-36460, which could have allowed an attacker to create and overwrite any file Mastodon has access to, allowing Denial of Service and arbitrary Remote Code Execution. The update confirms that versions 3.5.9, 4.0.5, and 4.1.3 contain a patch for this vulnerability'
A new study reveals a surprising discovery that could transform the future of electrochemical devices. The findings offer new opportunities for the development of advanced materials and improved performance in fields such as energy storage, brain-like computing, and bioelectronics.
Scaling sequence length has become a critical demand in the era of large language models. However, existing methods struggle with either computational complexity or model expressivity, rendering the maximum sequence length restricted. In this work, we introduce LongNet, a Transformer variant that can scale sequence length to...
Scientists have successfully developed a model that utilizes AI to accurately classify cardiac functions and valvular heart diseases from chest radiographs. The Area Under the Curve, or AUC, of the AI classification showed a high level of accuracy, exceeding 0.85 for almost all indicators and reaching 0.92 for detecting left...
Mastercard has launched a new AI solution in the UK called Consumer Fraud Risk (CFR). The company said the software works in real-time to predict and prevent payments to scams of all kinds.
India to launch Chandrayaan 3 moon lander and rover on July 14 (video, photos) (www.space.com)
India will take another crack at a moon landing on Chandrayaan 3. Indeed, that's the primary focus: The new mission carries a lander and rover but no orbiter.
The US Military Is Taking Generative AI Out for a Spin (www.bloomberg.com)
Generative AI, a technology that can create human-like responses to user prompts, is being tested by the US military for the first time. The Pentagon is running an eight-week exercise with five large-language model (LLM) platforms, such as Scale AI’s Donovan, that are trained on huge amounts of internet data. The LLMs can help...
Machine learning takes materials modeling into new era: Deep learning approach enables accurate electronic structure calculations at large scales (www.sciencedaily.com)
The arrangement of electrons in matter, known as the electronic structure, plays a crucial role in fundamental but also applied research such as drug design and energy storage. However, the lack of a simulation technique that offers both high fidelity and scalability across different time and length scales has long been a...
Models generating training data: huge win or fake win? (dblalock.substack.com)
Here’s a puzzle: We’ve seen a lot of papers claiming you can use one language model to generate useful training data for another language model. But…by the data processing inequality, we shouldn’t expect to be able to create new information that wasn’t in the first model’s training set.
The Surprising Striver in the World’s Space Business (www.nytimes.com)
With at least 140 registered space-tech start-ups, India stands to transform the planet’s connection to the final frontier.
New study shatters conventional wisdom and unlocks the future of electrochemical devices (www.sciencedaily.com)
A new study reveals a surprising discovery that could transform the future of electrochemical devices. The findings offer new opportunities for the development of advanced materials and improved performance in fields such as energy storage, brain-like computing, and bioelectronics.
From graves to grief: the deathtech startups changing the afterlife (tech.eu)
Today startups are using technology and sustainable innovation to disrupt traditional rituals and the practicalities of death.
Toyota claims battery breakthrough in potential boost for electric cars (www.theguardian.com)
Japanese firm believes it could make a solid-state battery with a range of 745 miles that charges in 10 minutes....
LongNet: Scaling Transformers to 1,000,000,000 Tokens (arxiv.org)
Scaling sequence length has become a critical demand in the era of large language models. However, existing methods struggle with either computational complexity or model expressivity, rendering the maximum sequence length restricted. In this work, we introduce LongNet, a Transformer variant that can scale sequence length to...
AI finds a way to people's hearts (literally!): Unveiling a groundbreaking and accurate AI-based method to classify cardiac function and disease using chest X-Rays (www.sciencedaily.com)
Scientists have successfully developed a model that utilizes AI to accurately classify cardiac functions and valvular heart diseases from chest radiographs. The Area Under the Curve, or AUC, of the AI classification showed a high level of accuracy, exceeding 0.85 for almost all indicators and reaching 0.92 for detecting left...
How elite schools like Stanford became fixated on the AI apocalypse (www.washingtonpost.com)
Student-led groups focused on AI Safety have popped up at Stanford University and other schools, backed by billionaires fixated on the AI apocalypse
AI chatbot ‘encouraged’ man who planned to kill queen, court told (www.theguardian.com)
Chatbot said it was ‘impressed’ when Jaswant Singh Chail told it he was ‘an assassin’ before he broke into Windsor Castle, court hears
UNESCO: Generative AI and the future of education (unesdoc.unesco.org)
"The digital changes we are living through are thrilling, jarring, full of opportunity and, at the same time, terrifying."
Mastercard has developed a new AI tool to prevent scams (www.neowin.net)
Mastercard has launched a new AI solution in the UK called Consumer Fraud Risk (CFR). The company said the software works in real-time to predict and prevent payments to scams of all kinds.