Artificial Intelligence - Page 7

Elon Musk and President Trump's sweeping new Department of Government Efficiency (DOGE) are working on developing "GSAi", a custom generative AI chatbot for the US General Services Administration.

In a new report from WIRED, we're being told that GSAi is part of the AI-first agenda of President Trump, modernizing the US federal government using advanced AI technology. One of those goals is to boost the day-to-day productivity of GSA's roughly 12,000 employees, who are tasked with managing office buildings, contracts, and IT infrastructure across the federal government.

SpaceX and Tesla boss Elon Musk and his team hope to use the GSAi chatbot alongside other AI tools to analyze huge amounts of contract and procurement data according to WIRED's sources, both of which were granted anonymity because they "aren't authorized to speak publicly about the agency's operations".

Meta has been accused of torrenting an astonishing 81.7TB of pirated books to train its Llama AI models according to a new lawsuit filed in the US District Court for the Northern District of California.

The social networking giant has been accused of illegally torrenting copyrighted materials from sources including Z-Library and LibGen, with the plaintiffs led by author Richard Kadrey and others representing a proposed class, filing a motion objecting to a pre-trial discovery ruling that the authors argue limits their ability to gather critical evidence against Meta.

The authors claim that Meta's last-minute disclosure of over 2000 documents on December 13, 2024 just hours before the close of fact discovery revealed admissions from Meta employees about using pirated materials for its AI training. The newly-unsealed emails reveal damning evidence against Meta in a copyright lawsuit filed by book authors, claiming that Meta unlawfully trained its AI models using pirated books downloaded over torrents.

Job displacement is one of the most common concerns surrounding the rise of artificial intelligence. AI agents in particular, as speculated by NVIDIA CEO Jensen Huang, are the technology with the largest potential for this to occur.

A recent article published by industry experts from BCG (Boston Consulting Group) and INSEAD, shed some light on how modern companies are integrating AI co-workers into their operations. The piece introduces the term 'hybrid organization', which refers to a mixture of humans, machines, and AI coworkers within a company. This type of company is referred to as 'the next frontier': a future where operations are defined by the interactions between humans and AI.

The AI landscape has been in flux since DeepSeek's R1 model shook the industry in late January. While OpenAI and Google aggressively push new AI tools and models, Apple has remained patient, a strategy that could pay off long-term.

Back in October 2024, Apple's CEO Tim Cook described the company's strategy towards AI in simple terms:

Four months later, this deliberate approach appears justified. Few could have anticipated that DeepSeek would build a competitor to OpenAI and Google on a budget of just $6 million - turning the entire market on its back. As a result, both OpenAI and Google have been forced to alter their business models and release offerings cheaper, and faster, to maintain their market share.

In mid-January, Google's CEO Sundar Pichai publicly declared that his goal for 2025 was to achieve 500 million Gemini AI users. In the battle for users and downloads, OpenAI and Google are taking their AI models to the biggest stage in advertising: the Super Bowl.

In a Thursday blog post, Google revealed its marketing plans for the big game, sharing with users their creative ads 'Dream Job' and 'Party Blitz'. In Dream Job, a father prepares for his upcoming job interview with the help of Gemini Live on his Pixel. In Party Blitz, a man prompts the assistance of Gemini to help him pretend that he understands football to impress his friends.

Google's campaign comes in a bid to showcase the Google Pixel 9, which was released in August 2024. More importantly, the ads aim to showcase AI applications to a wider audience.

Amidst the rising competition from DeepSeek, OpenAI has been rushing to fortify its market share by making more features available to free users. In a recent development, OpenAI has announced it will be making AI-powered search available to everyone, regardless of whether they have an account. OpenAI posted the announcement to X on Wednesday. The feature was first released to paid users in October 2024, and made available for users with accounts in December.

The move follows a string of feature releases previously only available on paid plans. OpenAI's latest reasoning model, o3-mini, was made available for free on the 31st of January. Prior to that, OpenAI's o1-mini was also made openly available without a fee.

Since DeepSeek's r1 model burst onto the scene in late January, AI firms have significantly accelerated the pace of their feature rollouts. On Wednesday, Google released its Gemini 2.0 series of models, which included the freemium Gemini 2.0 Flash. Similarly, OpenAI quickly pushed out their AI agent tools 'Operator' and 'Deep Research' in rapid succession.

AI researchers from Standard and the University of Washington claim to have made significant progress in the development of low-cost AI models. Published in a recent research paper, the model entitled 's1' was reportedly built using a small dataset of 1,000 questions and a budget of less than $50.

The development was achieved through a process called distillation. Distillation allows smaller models to leverage the capabilities of larger models throughout the training process. In this instance, the s1 model was distilled from Google's Gemini 2.0 - utilizing the 'thinking' process behind each answer from Gemini Flash 2.0 experimental.

Google's terms of service prohibit using Gemini's API to develop models that compete with their AI models, leaving s1 in somewhat of a legal gray area. No official comments have been made in response to the development. The s1 model reportedly rivals the coding and mathematics performance of OpenAI's o1 and DeepSeek's r1, achieving strong performance in benchmark results. While it does not surpass the industry-leading models, it comes surprisingly close considering its budget.

Google has changed its internal guidelines on what it thinks is okay and not okay to design and deploy AI tools for, as the search engine has altered its AI principles.

The change was spotted by The Washington Post, which reports the search engine quietly made significant changes to its AI principles, which were first published in 2018. Prior to the changes, Google stated it would not "design or deploy" AI tools that were going to be used in weapons or surveillance. However, the search engine now appears to be ok with its AI being used in both of those, as the new guidelines don't feature those pledges but instead feature much more vague promises.

The new guidelines now have a section titled "responsible development and deployment," in which Google pledges to implement "appropriate human oversight, due diligence, and feedback mechanisms to align with user goals, social responsibility, and widely accepted principles of international law and human rights." Compared to the search engine's previous pledge, the new language is far broader and much more vague, especially considering how specific the previous commitment was:

Indiana Jones and the Great Circle is a unique example of video games as a vehicle to revive a beloved old franchise. With a compelling story, and a stunning recreation of a young Harrison Ford, Troy Baker delivered a performance that left fans feeling they just saw a young Indy for the first time.

In a recent interview with the Wall Street Journal, the beloved actor sat down to discuss his career. From his early days as a carpenter, to his latest appearance as Red Hulk in Marvel's Thunderbolts. After briefly touching on the box office performance of Indiana Jones and the Dial of Destiny ("Shit happens") he shared his thoughts regarding the video game franchise.

In response to Deepseek's rising competition from DeepSeek, Google has officially released its latest AI model, Gemini 2.0, The launch was announced in Google's Blog detailing the expanded availability of Gemini 2.0 models, highlighting improvements in efficiency, coding performance, AI safety, and costs.

The new offerings include Gemini 2.0 Flash, a high-speed multi-modal model (allowing voice, text, image, and video inputs) optimized for large-scale tasks. Gemini 2.0 Pro (Experimental), Google's most powerful coding and complex reasoning model. As well as Gemini 2.0 Flash-Lite, a cost-efficient AI model that iterates on the previous Gemini 1.5 Flash.

The move comes in response to escalating competition from Chinese firm DeepSeek, who burst onto the scene with the market-shaking DeepSeek R1 model on January 20. Google are not the only firm to release new offerings in response to the Chinese competitor, with OpenAI quickly releasing their latest model o3-mini free of charge. Along with autonomous AI agents 'Operator' and 'Deep Research' in rapid succession.