Your Tech Story

AI

Large AI Dataset Has Over 1,000 Child Abuse Images, Researchers Find

Large AI Dataset Has Over 1,000 Child Abuse Images, Researchers Find

Leave a Comment / Artificial Intellegence, News, Tech News / By

A recent report by the Stanford Internet Observatory has uncovered a disturbing reality lurking within the foundation of popular artificial intelligence image-generators. The investigation reveals the presence of over 3,200 images suspected to be related to child sexual abuse within the LAION database, a colossal index of online images and captions utilized to train prominent AI image-making systems like Stable Diffusion.

Alarming Implications for AI-generated Content

Large AI Dataset Has Over 1,000 Child Abuse Images, Researchers Find

Image Source: borneobulletin.com.bn

This discovery sheds light on a concerning flaw in AI technology, indicating that the incorporation of such illicit images into AI datasets has facilitated the generation of realistic and explicit imagery featuring fake children. Moreover, it has enabled the transformation of innocuous social media pictures of clothed adolescents into explicit nudes, causing widespread alarm among educational institutions and law enforcement agencies globally.

Immediate Action and Ongoing Concerns

In response to the Stanford Internet Observatory’s revelations, LAION swiftly announced the temporary removal of its datasets, emphasizing a “zero tolerance policy for illegal content.” Despite constituting a fraction of LAION’s extensive image repository, these images significantly impact the capability of AI tools to generate harmful outputs while perpetuating the prior abuse suffered by the real victims depicted multiple times.

Challenges in Remediation and Accountability

David Thiel, chief technologist at the Stanford Internet Observatory, highlighted the challenges in rectifying this issue. He attributed the problem to the hurried deployment of numerous generative AI projects into the market, emphasizing the need for more rigorous scrutiny before open-sourcing vast datasets scraped from the internet.

Also Read: How to Install or Update the WiFi Driver on Windows 11

Unforeseen Ramifications and Industry Accountability

London-based startup Stability AI, a significant contributor to LAION’s dataset development, faces scrutiny for its role in shaping these datasets. While newer versions of their technology mitigate the creation of harmful content, an older, problematic version remains embedded in various applications. Lloyd Richardson from the Canadian Centre for Child Protection expressed concern over the prevalence of this older model and its widespread accessibility, acknowledging the difficulty in retracting it from multiple local machines.

The findings underscore the urgent need for greater responsibility and stringent measures within the AI development sphere to prevent the inadvertent perpetuation of exploitative content and to protect vulnerable individuals.

Former Pakistan PM Imran Khan Has Used AI Voice Clone to Campaign From Jail

Former Pakistan PM Imran Khan Has Used AI Voice Clone to Campaign From Jail

Leave a Comment / Artificial Intellegence, News, Tech News / By

Pakistan’s political landscape witnessed a groundbreaking event as former Prime Minister Imran Khan employed artificial intelligence to deliver a compelling speech to his supporters while incarcerated. This innovative move, leveraging AI to transcend physical constraints, has set a precedent in political campaigning.

Former Pakistan PM Imran Khan Has Used AI Voice Clone to Campaign From Jail

Image Source: insaf.pk

Imran Khan’s voice, replicated by AI, resonated across social media platforms during a virtual event, captivating over a million viewers with a meticulously crafted four-minute address. Despite Khan’s limited public access due to imprisonment, this technological breakthrough allowed him to communicate his message effectively, potentially influencing the political dynamics leading up to the forthcoming elections.

The utilization of AI to project Khan’s voice highlights the technologically progressive stance of his political party, PTI. This move not only underscores Khan’s enduring popularity but also positions PTI as an innovative force, distinguishing it from traditional political entities.

Implications for Pakistan's Political Future

While PTI confirmed the AI-generated nature of the speech, questions loom regarding its adherence to legal norms. Khan’s confinement and legal battles cast uncertainty on his ability to participate directly in the imminent parliamentary polls. Despite his lawyer’s mention of the potential submission of nomination papers pending a court decision, the legality of AI-delivered political addresses remains a point of contention.

Analysts speculate on the impact of this AI-driven approach. While acknowledging its potential to galvanize PTI’s support base, concerns persist over potential legal violations barring any direct or indirect public addresses by an incarcerated individual.

Legal Implications and Political Ramifications

The overwhelming response to Khan’s AI-mediated speech signals a pivotal shift in political campaigning, leveraging technological advancements to circumvent physical constraints. However, the government’s silence on this issue raises questions about the official response to such novel campaigning methods.

As the nation prepares for elections, the intersection of technology and politics introduces both excitement and legal ambiguity. The sheer scale of social media viewership juxtaposed against the broader voter base underscores the evolving landscape where technological outreach might significantly impact political outcomes.

Also Read: How To Fix If Your Samsung Tablet Is Running Slow?

Imran Khan’s AI-powered address has not only captivated the public but also sparked debates on the permissible extent of political engagement for incarcerated individuals. Whether this technological leap becomes a catalyst for change or raises legal hurdles remains to be seen, yet it undeniably heralds a new era in Pakistan’s political narrative.

Essential AI Comes Out of Stealth With $57 Million in Funding

Essential AI Comes Out of Stealth With $57 Million in Funding

Leave a Comment / Artificial Intellegence, News, Tech News / By

It has been claimed that Essential AI has come out of hiding with $56.5 million in fresh investment.

According to a Bloomberg News story published on Tuesday, December 12, the artificial intelligence (AI) business, which was started by two former Google employees, has created a technology known as “Enterprise Brain” that can be used for corporate duties including data analysis as well as automate tedious operations.

Essential AI Comes Out of Stealth With $57 Million in Funding

Image Source: bloomberg.com

The story claims that Niki Parmar and the chief executive officer Ashish Vaswani launched Essential. The two collaborated with other AI heavyweights at Google to write the “seminal” piece “Attention Is All You Need.”

The article said the study established the foundation for big language models, which power chatbots with artificial intelligence such as ChatGPT. According to earlier estimates, Essential had raised a total of $40 million.

Focus on LLMs

As mentioned back in August, LLMs have advanced artificial intelligence to new heights by allowing it to work with audio, pictures, video, plus music in addition to text.

“As they build, companies developing LLMs will contend with the challenges of collecting and classifying large amounts of data — as well as understanding the intricacies of how models now operate and how that differs from the previous status quo,” PYMNTS wrote.

pymnts.com

In the process of investing in LLMs as well as developing alliances, tech behemoths such as Alphabet along with Microsoft, as well as financiers such as Fusion Fund or Scale VC, have taken on a significant task which is ensuring that their LLM protégés gather and train on massive amounts of information to mould them so that they carry out and produce the intended outcomes.

The paper points out that data is useless on its own. It must be sorted, labelled, measured, grouped, and categorised in a variety of ways for models to make use of it. Labelling and classification data can also convey the correct meaning and context, conveying what an actual user planned to say.

Also Read: Two Ways to Reset Your Forgotten Apple Watch Passcode

“Steering these volumes of data through rule sets with correct context is a work in progress.” PYMNTS wrote. “The effort requires that the model reviews and connects the dots with whatever happened earlier or happens later in the chat or text.”

pymnts.com

The goal of Essential AI, a 2023 startup headquartered in San Francisco, California, is to strengthen the bond between human beings and technology by enabling cooperative capacities that go well beyond the scope of what is possible for humans to do on their own. Essential AI is creating full-stack AI solutions that swiftly learn to boost productivity by automating laborious and lengthy activities to completely transform the way businesses operate.

US Navy, UK, Australia to Test AI System to Help Crews Track Chinese Submarines in Pacific

US Navy, UK, Australia to Test AI System to Help Crews Track Chinese Submarines in Pacific

Leave a Comment / Artificial Intellegence, News, Tech News / By

The United States and two of its closest companions are going to test an emerging artificial intelligence-based method for tracking Chinese submarines.

On Friday, the defence leaders of the United States, the United Kingdom, and Australia said that crews operating the US Navy’s finest naval surveillance as well as battle aircraft on Pacific operations will leverage artificial intelligence (AI) algorithms to quickly evaluate sonar data collected by submerged devices from all three countries.

US Navy, UK, Australia to Test AI System to Help Crews Track Chinese Submarines in Pacific

Image Source: business-standard.com

As they look for measures to counteract the effects of China’s fast military modernization and rising global authority, the allies may be able to detect Chinese submarines more quickly and accurately thanks to this technology. The experiments are a component of Aukus Pillar II, a comprehensive technology cooperation agreement involving the three countries.

“These joint advances will allow for timely high-volume data exploitation, improving our anti-submarine warfare capabilities,” according to a joint statement from US Secretary of Defence Lloyd Austin, Australian Defence Minister Richard Marles and UK Secretary of State for Defence Grant Shapps during a meeting in California.

scmp.com

The three major powers said that they will use cutting-edge algorithms based on artificial intelligence (AI) on a variety of platforms, such as the P-8A Poseidon aircraft, to interpret data from each country’s sonobuoys, or underwater detecting sensors.

The Boeing Company’s maritime aircraft is flown by all three countries. The warplane, which is owned by the United States, is often patrolling the Pacific, such as the South China Sea, when Chinese fighters may sometimes swarm them. Poseidon is equipped with cruise missiles along with torpedoes to take out surface ships and submarines.

Aukus, a Greater Security Collaboration

The United States has sought a variety of regional partnerships to contain China, and the declarations are a component of the Aukus, a greater security collaboration involving the three allies.

The partnership’s first pillar, dedicated to enhancing Australia’s submarine capacity which is powered by nuclear energy, is to jointly construct a new submarine that will be deployed by 2040. Eight technology domains, such as quantum technologies, sophisticated cybersecurity, as well as hypersonic weaponry, are the focus of Pillar II’s collaboration efforts.

Also Read: Many PlayStation Users’ Accounts Locked Out, Get Permanent Suspension Message From Sony

According to the declaration, the three heads of defence will combine their capabilities to launch and retrieve underwater drone automobiles from torpedo tubes on their present submarines for undersea combat and information collection.

“This capability increases the range and capability of our undersea forces and will also support” Australia’s eventual new submarine called “SSN-Aukus,” the announcement read.

scmp.com
Amazon Has Developed Its Own Version of an AI Image Generator

Amazon Has Developed Its Own Version of an AI Image Generator

Leave a Comment / Artificial Intellegence, News, Tech News / By

In the bustling realm of technological advancements, Amazon has leaped into the burgeoning landscape of AI-powered image generation with its latest offering, the Titan Image Generator. The unveiling took place at Amazon’s AWS re:Invent conference, marking a strategic foray into the burgeoning domain of creative AI tools.

Revolutionizing Image Creation

The Titan Image Generator, currently available for preview on the Bedrock console for AWS customers, introduces a transformative approach to image creation. Users can either input textual prompts to generate images from scratch or upload existing images for editing. Amazon asserts that this tool has the capability to churn out high volumes of studio-quality, true-to-life images at an economical cost.

Amazon Has Developed Its Own Version of an AI Image Generator

Image Source: engadget.com

According to Amazon, the Titan Image Generator excels in producing contextually relevant images from intricate textual cues while ensuring precise object composition and minimal distortions. Moreover, the tool enables users to manipulate images effortlessly by isolating specific areas for modifications, such as background replacement or object substitution. The Generative Expand feature akin to Photoshop extends image boundaries by adding artificial elements.

One of the standout features of this AI-powered innovation is its incorporation of an imperceptible watermark onto generated images. Amazon purports that this measure will aid in curbing misinformation by discreetly identifying AI-generated content, fostering a secure and transparent landscape for AI technology development. Impressively, the watermarks are designed to resist alterations, fortifying their authenticity.

Beyond Image Generation

Not limited to image creation alone, the Titan Image Generator offers an extra layer of functionality by generating descriptive text, facilitating the seamless creation of social media posts or captions.

Also Read: Marshall Motif II ANC Review: Noise-Canceling Earbuds That Rock Your World

Amazon’s latest unveiling at the AWS re:Invent conference was not solely confined to the Titan Image Generator. The tech giant also showcased its cutting-edge AI chips and unveiled Q, a business-centric AI chatbot. Moreover, Amazon has recently rolled out a tool catering to advertisers, enabling them to incorporate AI-generated backgrounds into product images, further underscoring the company’s commitment to innovative AI-driven solutions.

With the debut of the Titan Image Generator, Amazon has positioned itself as a frontrunner in the rapidly evolving landscape of AI-powered image generation, heralding a new era of creativity and authenticity while addressing concerns related to misinformation and content transparency.

How to Detect AI-Generated Deepfake Videos?

How to Detect AI-Generated Deepfake Videos?

Leave a Comment / Tech News / By

In an era where technology is advancing rapidly, the emergence of deepfake videos has become a significant concern. These AI-generated clips, which can be manipulated and create realistic video content, pose serious challenges to individuals and society. In this article, we delve into the world of deepfakes and guide you on how to spot these sophisticated fakes.

The Rise of AI in Video Manipulation

How to Detect AI-Generated Deepfake Videos?

Image Source: euractiv.com

Artificial intelligence (AI) has transformed video editing, making it possible to create deepfakes that are increasingly difficult to distinguish from real footage. Understanding the mechanics behind the role of AI in video manipulation is the first step in learning how to spot these frauds.

Understanding How Deepfakes Are Made

Deepfakes are created using advanced AI techniques, primarily deep learning and neural networks. By analyzing and copying patterns in existing video footage, these AI systems can generate new content where individuals appear to say or do something they never actually did.

Recognizing Deepfakes

Detecting deepfakes involves examining various aspects of a video, from visual cues to audio synchronization.

Visual Inconsistencies

One key to detecting deepfakes is to look for visual anomalies. This includes unusual skin tones, poor lighting, or mismatched shadows that don’t correspond to natural settings.

Audio Analysis

Analyzing the audio may provide clues. A mismatch between spoken words and lip movements is a red flag, indicating possible manipulation.

Analyzing Facial Expressions

Deepfakes often struggle to accurately replicate natural facial expressions. Any unnatural or exaggerated expressions may indicate that a video is a deepfake.

Blinking Patterns

A subtle but obvious signal is the blinking pattern. AI-generated videos often fail to capture natural blinking, resulting in either excessive or insufficient blinking.

Technological Tools to Detect Deepfakes

As deepfakes become more sophisticated, technology to detect them evolves.

AI Detection Algorithms

AI algorithms are being developed to fight fire with fire. These algorithms can analyze videos for anomalies at a much finer level than the human eye.

Software Solutions

Various software tools are available that use AI and machine learning to flag deepfakes. These tools analyze the video frame by frame and look for obvious signs of manipulation.

The Role of Media Literacy

In the fight against deepfakes, media literacy plays an important role. It is essential to be a critical viewer and question the authenticity of questionable content.

Educating the Public

Educational initiatives about deepfakes can empower the public to identify and question manipulated content, thereby reducing the impact of these videos.

Legal and Ethical Implications

Deepfakes raise significant legal and ethical concerns ranging from defamation to misinformation. It is important to understand these implications in the broader context of dealing with deepfake technology.

Also Read: Tinder Introduces New Features To Keep Up With Changing Dating Dynamics

Conclusion

Detecting AI-generated deepfake videos requires a combination of visual and audio analysis, technical tools, and a critical approach to media consumption. As technology advances, it is important to remain informed and alert to identify these sophisticated counterfeits.

FAQs

  • What is the most common sign of a deepfake video? Look for visual inconsistencies and unnatural facial expressions, which are common indicators.

  • Can AI always detect deepfakes? Although AI is a powerful tool, it is not foolproof. Both deepfake and detection technologies continue to advance.

  • How can I improve my ability to spot deepfakes? Increase your media literacy and stay updated on the latest trends in deepfake technology.

  • Are there any legal measures to combat deepfakes? Legal measures are evolving, but vary by country and the specific use of deepfake technology.

  • Can deepfakes be harmless? While some deepfakes are created for entertainment, the potential for harm, especially in spreading misinformation, is significant.