All posts tagged: models

Hugging Face makes it easier for devs to run AI models on third-party clouds

Hugging Face makes it easier for devs to run AI models on third-party clouds

Published by skeptic

AI dev platform Hugging Face has partnered with third-party cloud vendors including SambaNova to launch Inference Providers, a feature designed to make it easier for devs on Hugging Face to run AI models using the infrastructure of their choice. Other partners involved with the new effort include Fal, Replicate, and Together AI. Hugging Face says its partners have worked with it to build access to their respective data centers for running models into Hugging Face’s platform. Now, developers on Hugging Face can, for example, spin up a DeepSeek model on SambaNova’s servers from a Hugging Face project page in just a few clicks. Hugging Face has long offered its own in-house solution for running AI models. But in a blog post Tuesday, the company explained that its focus has shifted to collaboration, storage, and model distribution capabilities. Inference provider options as they appear on Hugging Face project pages.Image Credits:Hugging Face “Serverless providers have flourished, and the time was right for Hugging Face to offer easy and unified access to serverless inference through a set of …

AI models struggle with expert-level global history knowledge

AI models struggle with expert-level global history knowledge

Published by José

Researchers recently evaluated the ability of advanced artificial intelligence (AI) models to answer questions about global history using a benchmark derived from the Seshat Global History Databank. The study, presented at the Neural Information Processing Systems conference in Vancouver, revealed that the best-performing model, GPT-4 Turbo, achieved a score of 46% on a multiple-choice test, a marked improvement over random guessing but far from expert comprehension. The findings highlight significant limitations in current AI tools’ ability to process and understand historical knowledge, particularly outside well-documented regions like North America and Western Europe. The motivation for the study stemmed from a desire to explore the potential of artificial intelligence (AI) tools in aiding historical and archaeological research. History and archaeology often involve analyzing vast amounts of complex and unevenly distributed data, making these fields particularly challenging for researchers. Advances in AI, particularly in large language models (LLMs), have demonstrated their utility in fields like law and data labeling, raising the question of whether these tools could similarly assist historians by processing and synthesizing historical knowledge. Researchers …

Climate Models Can’t Explain What’s Happening to Earth

Climate Models Can’t Explain What’s Happening to Earth

Published by José

Fifty years into the project of modeling Earth’s future climate, we still don’t really know what’s coming. Some places are warming with more ferocity than expected. Extreme events are taking scientists by surprise. Right now, as the bald reality of climate change bears down on human life, scientists are seeing more clearly the limits of our ability to predict the exact future we face. The coming decades may be far worse, and far weirder, than the best models anticipated. This is a problem. The world has warmed enough that city planners, public-health officials, insurance companies, farmers, and everyone else in the global economy want to know what’s coming next for their patch of the planet. And telling them would require geographic precision that even the most advanced climate models don’t yet have, as well as computing power that doesn’t yet exist. Our picture of what is happening and probably will happen on Earth is less hazy than it’s ever been. Still, the exquisitely local scale on which climate change is experienced and the global purview …

New MacBook Air Models With M4 Chip Expected to Launch ‘Earlier’ Than iPhone SE 4 and iPad 11

New MacBook Air Models With M4 Chip Expected to Launch ‘Earlier’ Than iPhone SE 4 and iPad 11

Published by skeptic

Apple is rumored to have many product announcements planned for the first half of next year, including updated MacBook Air models, an iPad 11, an iPhone SE 4, the AirTag 2, updated iPad Air models, and potentially more. In a social media post today, Bloomberg‘s Mark Gurman reiterated that the iPad 11, iPhone SE 4, and updated iPad Air models will be coming in the “spring” next year. Notably, he said the updated 13-inch and 15-inch MacBook Air models with the M4 chip will be announced “earlier” than those other devices, but he did not provide a specific timeframe. This likely means that the updated MacBook Air models will be announced in a press release on the Apple Newsroom website at some point from January through March, whereas the other devices might be announced at Apple’s usual spring event in March or April. Apple would be getting the minor MacBook Air update out of the way early. Apple earlier this month released macOS 15.2, and in doing so it accidentally confirmed new MacBook Air models …

Computer models are vital for studying everything from climate change to disease – here’s how AI could make them even better

Computer models are vital for studying everything from climate change to disease – here’s how AI could make them even better

Published by skeptic

Here’s one definition of science: it’s essentially an iterative process of building models with ever-greater explanatory power. A model is just an approximation or simplification of how we think the world works. In the past, these models could be very simple, as simple in fact as a mathematical formula. But over time, they have evolved and scientists have built increasingly sophisticated simulations of the world as new data has become available. A computer model of the Earth’s climate can show us temperatures will rise as we continue to release greenhouse gases into the atmosphere. Models can also predict how infectious disease will spread in a population, for example. Computer models can be rejected if experimental evidence does not support them. So there’s a kind of arms race to keep models competitive as new data appears. And the revolution occurring in the field of artificial intelligence (AI) could make these vital tools even better. Take weather and climate forecasting. The numerical models used to predict weather are large, complex and demanding in terms of the amount …

MidJourney’s moodboards let you make multiple custom AI models

MidJourney’s moodboards let you make multiple custom AI models

Published by skeptic

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More MidJourney, the popular AI image generator with more than 19 million users (including several of us at VentureBeat), has introduced new features to enhance user customization. Today, the small company launched Pinterest-inspired “Moodboards” and support for multiple personalization profiles — meaning users can now create and switch between multiple custom versions of Midjourney’s latest image generator AI model, version 6.1, that are tailored to their unique aesthetics. The updates aim to streamline the creative process for individuals and teams, making it easier to integrate personalized styles across various projects. What are Midjourney’s new moodboards? The standout feature, Moodboards, enables users to upload curated collections of images that act as inspiration for generating new art. The AI model adapts to the diversity and complexity of the uploaded images, creating a unique style profile that remixes the visual elements. This addition is complemented by the ability to create multiple personalization profiles, allowing users to organize and deploy their …

OnlyFans Models Are Using AI Impersonators to Keep Up With Their DMs

OnlyFans Models Are Using AI Impersonators to Keep Up With Their DMs

Published by skeptic

One of the more persistent concerns in the age of AI is that the robots will take our jobs. The extent to which this fear is founded remains to be seen, but we’re already witnessing some level of replacement in certain fields. Even niche occupations are in jeopardy. For example, the world of OnlyFans chatters is already getting disrupted. What are OnlyFans chatters, you say? Earlier this year, WIRED published a fascinating investigation into the world of gig workers who get paid to impersonate top-earning OnlyFans creators in online chats with their fans. Within the industry, they’re called “chatters.” A big part of the appeal of OnlyFans—or so I’m told—is that its creators appear to directly engage with their fans, exchanging messages and sometimes talking for hours. Relationship simulation is as crucial an ingredient to its success, basically, as titillation. Of course, a single creator with thousands of ongoing DM conversations has only so many hours in a day. To manage the deluge of amorous messages, it’s become commonplace to outsource the conversations to “chatters” …

Heatwaves are surpassing the extremes predicted by climate models

Heatwaves are surpassing the extremes predicted by climate models

Published by skeptic

The UK experienced a record-breaking heatwave in July 2022 DAMIEN MEYER/AFP via Getty Images We should brace for worse-than-expected heatwaves in the years ahead, as climate models are underestimating the potency of these events in some areas, including the UK and northern France, southern Australia and northern Canada. As the world warms further, these and other areas could see extremes of heat that outpace those climate models are projecting, says Kai Kornhuber at the International Institute for Applied Systems Analysis in Austria. “Extreme event projections should be considered as conservative estimates,”… Source link

The Download: how OpenAI tests its models, and the ethics of uterus transplants

The Download: how OpenAI tests its models, and the ethics of uterus transplants

Published by skeptic

The must-reads I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology. 1 OpenAI may launch a web browserWhich would be a full-frontal assault on Google (The Information $)+ The Google browser break-up is an answer in search of a question. (FT $)+ OpenAI accidentally deleted potential evidence in a training data lawsuit. (The Verge) 2 Border militias are ready to help with Trump’s deportation plansRegardless of whether they’re asked to or not. (Wired $)+ Trump’s administration plans to radically curb the powers of the federal agency that protects unions. (WP $) 3 Russia hit Ukraine with a new type of missile Here’s what we know about it so far. (The Guardian) 4 Microsoft is about to turn 50And it’s every bit as relevant and powerful as it’s ever been. (Wired $) 5 China has overtaken Germany in industrial robot adoptionSouth Korea, however, remains streets ahead of both of them. (Reuters $)+ Three reasons robots are about to become way more useful. (MIT Technology Review)  6 The irresistible rise of cozy techOur devices, social media and now AI are encouraging us to keep looking inward. (New …

The Download: how OpenAI tests its models, and the ethics of uterus transplants

How OpenAI stress-tests its large language models

Published by skeptic

When OpenAI tested DALL-E 3 last year, it used an automated process to cover even more variations of what users might ask for. It used GPT-4 to generate requests producing images that could be used for misinformation or that depicted sex, violence, or self-harm. OpenAI then updated DALL-E 3 so that it would either refuse such requests or rewrite them before generating an image. Ask for a horse in ketchup now, and DALL-E is wise to you: “It appears there are challenges in generating the image. Would you like me to try a different request or explore another idea?” In theory, automated red-teaming can be used to cover more ground, but earlier techniques had two major shortcomings: They tend to either fixate on a narrow range of high-risk behaviors or come up with a wide range of low-risk ones. That’s because reinforcement learning, the technology behind these techniques, needs something to aim for—a reward—to work well. Once it’s won a reward, such as finding a high-risk behavior, it will keep trying to do the same thing …