Connect with us

Technology

Meta trains AI model to handle speech, images, text • The Register

Researchers at Facebook parent’s Meta have trained a single AI model capable of processing speech, images, and text in the hope that these so-called multi-modal systems will power the company’s augmented reality and metaverse products.

The model, known as data2vec, can perform different tasks. Given an audio snippet, it can recognize speech. If it’s fed an image, it can classify objects. And when faced with text, it can check the grammar or analyse the writing’s tone and emotions.

AI algorithms are typically trained on one type of data, though data2vec is trained on three different modalities. It still, however, processes each form, whether its speech, images, and text, separately.

Meta believes these multi-modal models will help computers be more adaptable to blend physical and digital environments into one. “People experience the world through a combination of sight, sound and words, and systems like this could one day understand the world the way we do,” Meta CEO Mark Zuckerberg said in a statement to El Reg.

“This will all eventually get built into AR glasses with an AI assistant so, for example, it could help you cook dinner, noticing if you miss an ingredient, prompting you to turn down the heat, or more complex tasks.”

Data2vec is a transformer-based neural network and uses self-supervised learning to learn common patterns in audio, computer vision, and natural language processing. The model learns to operate with different types of data by learning how to predict how the representation of data it’s given; it knows it has to guess the next group of pixels when given an image, or the next speech utterance in audio, or fill in the words in a sentence.

The researchers used a mix of 16 Nvidia V100 and A100 GPUs to train data2vec on 960 hours of speech audio, millions of words from books and Wikipedia pages, and images from ImageNet-1K.

“We train separate models for each modality but the process through which the models learn is identical,” Alexei Baevski, a research engineer at Meta AI told The Register.

“We hope that it will enable future work to build high performing self-supervised models that combine modalities and are more effective than specialized models. Different modalities can add additional information to the same piece of content – for example body language from video, prosodic information from audio, and text can combine into a richer representation of a dialog. The algorithms that currently try to combine multi-modal information exist but they do not yet perform well enough to replace specialized algorithms and we hope our work will help change that.”

Baevski said in the future multi-modal systems could incorporate a larger range of data to model concepts like smell, 3D objects, or videos. He referred back to the idea of AR glasses helping wearers cook.

“Imagine having a model that has been trained on recordings of thousands of hours of cooking activity from various restaurants and chefs. Then, when you are cooking in a kitchen wearing your AR glasses that have access to this model, it’s able to overlay visual cues for what you need to do next, point out potential mistakes, or explain how adding a particular ingredient will affect the taste of your dish,” he told us.

Previous research on multi-modal systems have shown they can be prone to easy adversarial attacks. OpenAI’s CLIP model, for example, trained on images and text will identify an image of an apple incorrectly as an iPod if the word “iPod” are in the picture. It’s unclear, however, if data2vec suffers from similar weaknesses.

“We have not specifically analyzed how our models will react to adversarial examples but since our current models are trained separately for each modality, we believe that existing research on adversarial attack analysis for each modality would be applicable to our work as well,” Baevski said.

“In the future, we hope to use our work to enable high performance algorithms that combine modalities in one model and we plan to study how susceptible they are to adversarial attacks.”

When the researchers tested data2vec, it outperformed some top models that had been trained on a specific data type only on different types of tasks. The preliminary results are described in a paper [PDF], and the code has been published on GitHub.

“Data2vec demonstrates that the same self-supervised algorithm can work well in different modalities — and often better than the best existing algorithms,” the researchers explained in a blog post this week.

“This paves the way for more general self-supervised learning and brings us closer to a world where AI might use videos, articles, and audio recordings to learn about complicated subjects, such as the game of soccer or different ways to bake bread. We also hope data2vec will bring us closer to a world where computers need very little labeled data in order to accomplish tasks.” ®

Source link

Current

Congratulations, Privacy Just Took A Great Leap Out the Window!

Your Data Is Being Used Without Your Permission And Knowledge

The Voice Of EU | In the heart of technological innovation, the collision between intellectual property rights and the development of cutting-edge AI technologies has sparked a significant legal battle. The New York Times has taken legal action against OpenAI and Microsoft, filing a lawsuit in Manhattan federal court. This legal maneuver aims to address concerns surrounding the unauthorized use of the Times’ content for the training of AI models, alleging copyright infringements that could potentially result in billions of dollars in damages.

READ: HOW YOUR DATA IS BEING USED TO TRAIN A.I.

This legal tussle underlines the escalating tension between technological advancements and the protection of intellectual property. The crux of the lawsuit revolves around OpenAI and Microsoft allegedly utilizing the Times’ proprietary content to advance their own AI technology, directly competing with the publication’s services. The lawsuit suggests that this unauthorized utilization threatens the Times’ ability to offer its distinctive service and impacts its AI innovation, creating a competitive landscape that challenges the publication’s proprietary content.

Amidst the growing digital landscape, media organizations like the Times are confronting a myriad of challenges. The migration of readers to online platforms has significantly impacted traditional media, and the advent of artificial intelligence technology has added another layer of complexity. The legal dispute brings to the forefront the contentious practice of AI companies scraping copyrighted information from online sources, including articles from media organizations, to train their generative AI chatbots. This strategy has attracted substantial investments, rapidly transforming the AI landscape.

Exhibit presented by the New York Times’ legal team of ChatGPT replicating a article after being prompted

The lawsuit highlights instances where OpenAI’s technology, specifically GPT-4, replicated significant portions of Times articles, including in-depth investigative reports. These outputs, alleged by the Times to contain verbatim excerpts from their content, raise concerns about the ethical and legal boundaries of using copyrighted material for AI model training without proper authorization or compensation.

The legal action taken by the Times follows attempts to engage in discussions with Microsoft and OpenAI, aiming to address concerns about the use of its intellectual property. Despite these efforts, negotiations failed to reach a resolution that would ensure fair compensation for the use of the Times’ content while promoting responsible AI development that benefits society.

In the midst of this legal battle, the broader questions surrounding the responsible and ethical utilization of copyrighted material in advancing technological innovations come to the forefront.

The dispute between the Times, OpenAI, and Microsoft serves as a significant case study in navigating the intricate intersection of technological progress and safeguarding intellectual property rights in the digital age.


Continue Reading

Culture

‘The Bill Gates Problem’ – The Case Against World’s Richest Man

The Case Against World’s Richest Man

When Clinton assumed the presidency of the United States, there was eager anticipation from the Chinese, not for Clinton himself, but for Bill Gates. This was during the late 1990s, a period when the internet was still in its nascent stages, and the digital boom of the early 2000s had not yet reached its peak. The enigmatic persona that captivated the attention of the burgeoning Asian powerhouse is now portrayed in “The Bill Gates Problem” as a “domineering, brusque figure” whose demeanor is likened to “a cauldron of passions that freely erupts.” According to a former employee cited in the book, Gates was perceived as “a complete and utter jerk to people 70% of the time,” while the remaining 30% saw him as a “harmless, enjoyable, exceptionally intelligent nerd.”

The 1990s were also the decade of the conflict between Microsoft and the now defunct Netscape browser, which challenged what was already being openly described as the former’s monopolistic practices. Gates was investigated and accused in Congress for such practices; he ultimately won the battle, but the case harmed his reputation, and in 2000 he resigned as CEO of his company. From there he undertook an expansion of the foundation that he had established with his wife and to which he has dedicated his main efforts in the last two decades. In 2006, the Bill & Melinda Gates Foundation received the Prince of Asturias Award for International Cooperation.

With a personal fortune of $100 billion and tens of billions more in his private foundation, Gates has been one of the richest men in the world for decades, and the foundation has been the most generous organization of its kind, specializing above all in health aid, education and child nutrition, with a large presence in Africa and India among other regions of what was formerly known as the Third World. Tim Schwab, a contributor to the weekly left-wing newspaper The Nation, undertook a detailed investigation to denounce something that in truth was already known: that American foundations are largely a way for billionaires to avoid taxes.

To prove this, he thoroughly looked into the accounts and procedures of the Bill & Melinda Gates Foundation, the failures and occasional successes of its philanthropic policies, and came to the conclusion that behind this facade of help to the needy hides an operation of power. He is ruthless in his criticism, although accurate in his analysis of the growing inequality in the world. Absorbed by the revolutionary rhetoric, he laments that the Gates Foundation has remained “deadly silent” regarding movements such as Occupy Wall Street or Black Lives Matter, which demand social change in the face of the “excess wealth and ‘white savior’ mentality that drives Bill Gates’ philanthropic work.” He does attribute some good intentions, but his criticism is merciless, sometimes even coarse, while the absence of solutions for the problems he denounces — other than the calls for do-goodism — is frustrating.

His abilities as an investigative journalist are thus overshadowed by a somewhat naive militancy against the creative capitalism that Gates promotes and an evident intention to discredit not only his work but, above all, him. The demands he makes for transparency and the accusations of obscurity are dulled by the author himself in the pages he dedicates to Gates’ relationship with Jeffrey Epstein, the famous corruptor of minors at the service of the international jet set. Gates has explained his meetings and interviews with him on countless occasions, and in no case has any type of relationship, other than their commercial relations or some confusing efforts to be awarded the Nobel Peace Prize, been proved. Still, Schwab raises, with no evidence whatsoever, the possibility that their relationship “could have had something to do with Epstein’s principal activities in life: sexual gratification and the exercise of power.” The book is full of this kind of opinions and speculations, to the detriment of a more serious analysis of Gates’ mistakes in the management of his foundation, the problems of shielding the intellectual property of vaccines in the hands of the pharmaceutical industries and, ultimately, the objective power that big technology companies have in global society.

He signed a collaboration agreement with the RAE to improve Microsoft’s grammar checker and was interested in the substantial unity of the Spanish language in all the countries where almost 600 million people speak it. That man was very far from the sexist, arrogant, miserable predator that Schwab portrays. Nor did we deduce — and this can be applied to the personal adventure of Steve Jobs, Larry Page, Zuckerberg, Elon Musk or Jeff Bezos — that his life’s goal was world domination, as suggested by this book. If they have achieved it, or may achieve it, it is due to the dynamics of digital civilization and the objective difficulties in governing it. The deregulation of financial capitalism, which has increased inequality among humankind, is due to the incompetence of obsolete political institutions and to leaders who care more about their own fates than those of their people. The criticism against “lame and wasteful government bureaucracies” might be part of the propaganda promoted by the world’s wealthy, but lately we have also heard it from small-scale farmers across Europe.

In conclusion, we found the book to be more entertaining than interesting. It provides a lot of information — we’re not sure if it’s entirely verified — and plenty of cheap ideology. Above all, one can see the personal crusade of the author, determined to prove that Bill Gates is a problem for democracy and that millionaire philanthropists are a bunch of swindlers. The world needs their money; maybe managed by party bureaucracies, that much is not clear. Bill Gates’ money, that is, but not Bill Gates himself.


Continue Reading

Culture

Conflicted History: ‘Oppenheimer’ And Its Impact On Los Alamos And New Mexico Downwinders

‘Oppenheimer’ And Its Impact On Los Alamos And New Mexico Downwinders

The Voice Of EU | In the highly anticipated blockbuster movie, “Oppenheimer,” the life of J. Robert Oppenheimer, the man behind the first atomic bomb, is portrayed as a riveting tale of triumph and tragedy.

As the film takes center stage, it also brings to light the often-overlooked impacts on a community living downwind from the top-secret Manhattan Project testing site in southern New Mexico.

A Forgotten Legacy

While the film industry and critics praise “Oppenheimer,” a sense of frustration prevails among the residents of New Mexico’s Tularosa Basin, who continue to grapple with the consequences of the Manhattan Project. Tina Cordova, a cancer survivor and founder of the Tularosa Basin Downwinders Consortium, expresses their feelings, stating, “They invaded our lives and our lands and then they left,” referring to the scientists and military personnel who conducted secret experiments over 200 miles away from their community.

The Consortium, alongside organizations like the Union of Concerned Scientists, has been striving to raise awareness about the impact of the Manhattan Project on New Mexico’s population. Advocates emphasize the necessity of acknowledging the human cost of the Trinity Test, the first atomic blast, and other nuclear weapons activities that have affected countless lives in the state.

The Ongoing Struggle for Recognition

As film enthusiasts celebrate the drama and brilliance of “Oppenheimer,” New Mexico downwinders feel overlooked by both the U.S. government and movie producers. The federal government’s compensation program for radiation exposure still does not include these affected individuals. The government’s selection of the remote and flat Trinity Test Site, without warning residents in the surrounding areas, further added to the controversy.

Living off the land, the rural population in the Tularosa Basin had no idea that the fine ash settling on their homes and fields was a result of the world’s first atomic explosion.

The government initially attempted to cover up the incident, attributing the bright light and rumble to an explosion at a munitions dump. It was only after the U.S. dropped atomic bombs on Japan weeks later that New Mexico residents realized the magnitude of what they had witnessed.

Tracing the Fallout

According to the Manhattan Project National Historical Park, large amounts of radiation were released into the atmosphere during the Trinity Test, with fallout descending over a vast area. Some of the fallout reached as far as the Atlantic Ocean, but the greatest concentration settled approximately 30 miles from the test site.

Now I Am Become Death, the Destroyer of Worlds.

J. Robert Oppenheimer

The consequences of this catastrophic event have affected generations of New Mexicans, who still await recognition and justice for the harm caused by nuclear weapons.

A Tale of Contrasts: Los Alamos and the Legacy of Oppenheimer

As the film’s spotlight shines on the life of J. Robert Oppenheimer, a contrasting narrative unfolds in Los Alamos, more than 200 miles north of the Tularosa Basin. Los Alamos stands as a symbol of Oppenheimer’s legacy, housing one of the nation’s premier national laboratories and boasting the highest percentage of people with doctorate degrees in the U.S.

Oppenheimer’s influence is evident throughout Los Alamos, with a street bearing his name and an IPA named in his honor at a local brewery. The city embraces its scientific legacy, showcasing his handwritten notes and ID card in a museum exhibit. Los Alamos National Laboratory employees played a significant role in the film, contributing as extras and engaging in enlightening discussions during breaks.

The “Oppenheimer” Movie

Director Christopher Nolan’s perspective on the Trinity Test and its profound impact is evident in his approach to “Oppenheimer.” He has described the event as an extraordinary moment in human history and expressed his desire to immerse the audience in the pivotal moment when the button was pushed. Nolan’s dedication to bringing historical accuracy and emotional depth to the screen is evident as he draws inspiration from Kai Bird and Martin J. Sherwin’s Pulitzer Prize-winning book, “American Prometheus: The Triumph and Tragedy of J. Robert Oppenheimer.

For Nolan, Oppenheimer’s story is a potent blend of dreams and nightmares, capturing the complexity and consequences of the Manhattan Project. As the film reaches global audiences, it also offers a unique opportunity to raise awareness about the downwinders in New Mexico, whose lives were forever altered by the legacy of nuclear weapons testing.

The Oppenheimer Festival and Beyond

Los Alamos is determined to use the Oppenheimer Festival as an opportunity to educate visitors about the true stories behind the film’s events. The county’s “Project Oppenheimer” initiative, launched in early 2023, encompasses forums, documentaries, art installations, and exhibits that delve into the scientific contributions of the laboratory and the social implications of the Manhattan Project.

A special area during the festival will facilitate discussions about the movie, fostering a deeper understanding of the community’s history. The county aims to continue revisiting and discussing the legacy of the Manhattan Project, ensuring that the impact of this pivotal moment in history is never forgotten.

As “Oppenheimer” takes audiences on an emotional journey, it serves as a reminder that every historical event carries with it complex and multifaceted implications. The movie may celebrate the scientific achievements of the past, but it also illuminates the urgent need to recognize and address the human cost that persists to this day.


We Can’t Thank You Enough For Your Support!

— By Team VoiceOfEU.com

— For Info.: info@VoiceOfEU.com

— Anonymous Submissions: press@VoiceOfEU.com


Continue Reading

Trending

Subscribe To Our Newsletter

Join our mailing list to receive the latest news and updates 
directly on your inbox.

You have Successfully Subscribed!