Lifeboat Foundation

NEW YORK, June 13 (Reuters) — Meta Platforms (META.O) said on Tuesday that it would provide researchers with access to components of a new “human-like” artificial intelligence model that it said can analyze and complete unfinished images more accurately than existing models.

The model, I-JEPA, uses background knowledge about the world to fill in missing pieces of images, rather than looking only at nearby pixels like other generative AI models, the company said.

That approach incorporates the kind of human-like reasoning advocated by Meta’s top AI scientist Yann LeCun and helps the technology to avoid errors that are common to AI-generated images, like hands with extra fingers, it said.

Starting today, users in the US can see how women’s tops from H&M, Anthropologie, Everlane, and Loft will look on a wide range of real human models. New filters aim to make finding clothes easier.

Google is introducing two new features for its online shopping experience that are designed to help users search for clothes in more detail and better visualize how clothing will look on different body types.

Starting today, Google Shopping users in the US can access a virtual try-on experience that realistically displays how an item of clothing will look on a selection of real human models. These models are available with various skin tones, ethnicities, hair types, and body shapes, ranging in size… More.

Continue reading “Google Shopping introduces new AI modeling feature” »

OpenAI CEO Sam Altman over the weekend called for enhanced collaboration between the U.S. and China on artificial intelligence development. Without mentioning the fact that his company’s products like ChatGPT are not available in China, he argued that China should be a major player in ensuring the safety of global AI development and rollout.

“With the emergence of the increasingly powerful AI systems, the stakes for global cooperation have never been higher,” he said in the keynote address for a conference hosted by the Beijing Academy of Artificial Intelligence, sounding more like someone leading an advocacy group on responsible tech than what he is: the CEO of a company responsible for shepherding that emergence.

Continue reading “OpenAI’s Sam Altman Makes Global Call For AI Regulation—And Includes China” »

A team of researchers in the United States and United Kingdom say they have created the world’s first synthetic human embryo-like structures from stem cells, bypassing the need for eggs and sperm.

These embryo-like structures are at the very earliest stages of human development: They don’t have a beating heart or a brain, for example. But scientists say they could one day help advance the understanding of genetic diseases or the causes of miscarriages.

The research raises critical legal and ethical questions, and many countries, including the US, don’t have laws governing the creation or treatment of synthetic embryos.

“Even though recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one—not even their creators—can understand, predict, or reliably control,” stated the letter.

The idea is that AI development should be “planned for and managed with commensurate care and resources.” However, the authors of the letter say that this level of planning is not happening. This leads to AI systems that are out of control.

Transparently communicating the limitations of the model and providing clear disclaimers when interacting with users can foster trust and accountability.

This paper demonstrates that a stochastic, rather than uniform, injection rate of MeV cosmic rays from supernovae can successfully model the observed ionization rate distribution for Galactic molecular clouds.

Qualcomm Technologies, Inc. unveiled the Qualcomm Video Collaboration Platform, a new suite of video collaboration solutions that allows original equipment manufacturers (OEMs) to easily design and deploy video conferencing products featuring superior video, audio and customizable on-device AI to power engaging, immersive virtual meeting experiences across enterprise, healthcare, educational, and home environments. The Qualcomm® Video Collaboration Platform is a one-stop solution that provides essential hardware and software features specifically tailored for video conferencing so that customers can quickly design and deploy a wide variety of video conferencing products, from enterprise video collaboration systems and huddle room systems to digital whiteboards, to touch controllers and personal devices for the home.

With support for Android and Linux, the three AI-rich platforms offer greater flexibility and ability to customize and deploy video conferencing products across diverse environments. Qualcomm Technologies’ industry-leading innovations in connectivity, compute, AI, audio, and video work together to deliver features that eliminate distractions, enhance productivity, and allow remote meeting callers to feel more connected to conference room participants by providing individual views of everybody in the room, creating an equal viewing experience for all participants.

With the rapid advances in generative AI, future meeting experiences will offer even more advanced video, speech, and text capabilities. Collaboration devices with dedicated hardware support for on-chip AI acceleration will be able to optimize these experiences by splitting workloads between the cloud and edge-based device.

AI isn’t programming’s future, it’s its present.

Scientists have created one of the most detailed 3D images of the synapse, the important juncture where neurons communicate with each other through an exchange of chemical signals. These nanometer-scale models will help scientists better understand and study neurodegenerative diseases such as Huntington’s disease and schizophrenia.

The new study appears in the journal PNAS and was authored by a team led by Steve Goldman, MD, Ph.D., co-director of the Center for Translational Neuromedicine at the University of Rochester and the University of Copenhagen. The findings represent a significant technical achievement that allows researchers to study the different cells that converge at individual synapses at a level of detail not previously achievable.

“It is one thing to understand the structure of the synapse from the literature, but it is another to see the precise geometry of interactions between individual cells with your own eyes,” said Abdellatif Benraiss, Ph.D., a research associate professor in the Center for Translational Neuromedicine and co-author of the study. “The ability to measure these extremely small environments is a young field, and holds the potential to advance our understanding of a number of neurodegenerative and neuropsychiatric diseases in which synaptic function is disturbed.”