AI Tools

1min.AI Comprehensive AI Suite

1min.AI provides a range of AI-driven tools including writing assistance for blog posts, background removal, text-to-speech conversion, and image upscaling, designed to streamline workflows for professionals, students, and creatives. The platform supports AI model chats with GPT-4 and Gemini Pro, along with audio/video editing, transcription, and translation. Users can earn free credits daily and accumulate rewards through reviews or referrals. A lifetime subscription is priced at £31.66 (discounted from £185.24).

Technology and Legal Disputes

Apple vs. Epic Games Antitrust Trial

The legal battle between Epic Games and Apple, ongoing since August due to Epic's addition of a direct payment mechanism in Fortnite violating Apple's rules, is heading to trial on May 3rd. Apple removed Fortnite from the App Store, leading Epic to file an antitrust lawsuit claiming the App Store is a monopoly. The trial is expected to reveal details about Apple's operations, with testimony from figures like Tim Cook and Tim Sweeney. The dispute centers on Apple's App Store policies, with Epic challenging the restrictions and Apple defending its control. The core issue is whether Apple's App Store policies are monopolistic and anti-competitive.

Artificial Intelligence

LLMs and RAG

Large Language Models (LLMs) excel in writing and answering questions but may lack accuracy. Retrieval-Augmented Generation (RAG) enhances LLMs by offering factual data, thus ensuring precision. LLMs generate human-like text but sometimes guess incorrectly. RAG retrieves relevant information from trusted sources, like books or databases, to provide accurate answers. RAG ensures the LLM stays accurate and truthful by grounding its responses in reliable information. LLMs create, and RAG monitors.

AI and Copyright Protest

A group of 1,000 musicians in the U.K. released a "silent album" titled “Is This What We Want?” to protest planned changes to copyright law that would make it easier to train AI on copyrighted work without licensing or payment. The album features recordings of empty studios and performance spaces from artists like Kate Bush, Max Richter, and Imogen Heap, with track titles spelling out "The British government must not legalize music theft to benefit AI companies." Organized by Ed Newton-Rex, the protest is part of a larger campaign against AI training without licensing, with a petition garnering over 47,000 signatures from creative professionals.

AI & Technology

Legal Challenges in AI

Chegg is suing Google, alleging that Google's AI-powered search summaries are causing a decline in their traffic and revenue. Chegg accuses Google of unfair competition, including reciprocal dealing, monopoly maintenance, and unjust enrichment. They are seeking compensatory damages and an injunction against Google's practices. This is not the first time a publisher has raised concerns about Google's AI integration into search. ![][https://techcrunch.com/wp-content/uploads/2024/01/GettyImages-1288924984.jpg?w=1390&crop=1]

AI Funding and Valuation

Anthropic is reportedly closing a $3.5 billion funding round, valuing the company at $61.5 billion. Investors including Lightspeed Venture Partners, General Catalyst, Bessemer Venture Partners, and MGX are expected to participate. This funding will support the development of more advanced AI technologies. Anthropic, known for its Claude AI chatbot, has reached approximately $1.2 billion in annualized revenue but remains unprofitable.

AI and National Security

Scale AI CEO Alexandr Wang is advocating for the U.S. to "win the AI war," viewing AI as a critical component of national security. He argues that AI could allow China to surpass Western military might. His concerns include the potential for Chinese LLM models to reflect communist viewpoints and the risk of government data gathering. Scale AI will also assist Qatar in developing 50 AI-powered government applications. Scale AI is a company known for its AI weapons and AI model training. ![][https://techcrunch.com/wp-content/uploads/2024/04/GettyImages-1152044635.jpg?resize=1200,800]

AI-Powered Weather Forecasting

The European Centre for Medium-Range Weather Forecasts (ECMWF) has launched an AI-powered forecasting model, AIFS, which it claims outperforms state-of-the-art physics-based models by up to 20%. AIFS operates faster and uses significantly less energy. AI-driven models like AIFS learn complex relationships from data. Google DeepMind's GenCast model has also demonstrated strong performance in AI-powered weather prediction. ECMWF states that AIFS complements its existing model.

Online Safety & Regulation

UK's Online Safety Act

Ofcom, the U.K.’s internet safety regulator, has released draft guidance to protect women and girls from online threats such as harassment, misogyny, and intimate image abuse, as part of the Online Safety Act (OSA) implementation. The government prioritizes the protection of women and girls, outlining specific forms of misogynistic abuse in the law. The OSA was approved in September 2023, faces criticism, and enforcement is expected to begin soon regarding illegal content and child protection. Full enforceability of the recommendations is projected for 2027 or later.

AI and Technology

Anthropic's New Model Excels in Coding

Anthropic has launched its first simulated reasoning model, demonstrating exceptional performance in coding tasks.

Politics & Technology

AI-Generated Video Incident at HUD

An AI-generated video showing Donald Trump bowing to and kissing Elon Musk's feet was displayed at the Department of Housing and Urban Development (HUD) headquarters. The video, played on a loop in the HUD cafeteria, was posted online by a Vox reporter. The incident occurred around the time of Musk's deadline for federal workers to report their accomplishments, triggering online trolling. A HUD spokesperson denounced the video as a misuse of taxpayer funds and resources, promising corrective action.

Artificial Intelligence

Anthropic's Claude 3.7 Sonnet Launch

Anthropic has introduced Claude 3.7 Sonnet, positioning it as a challenger to OpenAI and DeepSeek in the AI landscape. This new AI model is described as the first "hybrid reasoning model," capable of both quick responses and step-by-step problem-solving based on user preference. Claude 3.7 Sonnet emphasizes a unified approach to reasoning, contrasting with OpenAI's multiple models. It is optimized for real-world tasks, including improvements in coding and front-end web development, and is accessible across all Claude plans, including the free tier, and through the company's API, Amazon Bedrock, and Google Cloud's Vertex AI. Anthropic is also launching Claude Code, an "agentic" tool for delegating coding tasks directly from a terminal interface, currently available as a limited release.

AI Applications

AI in Traffic Safety

AI is being used to improve road safety by detecting distracted driving. Acusensus' "Heads Up" cameras use AI to identify phone use and seatbelt violations. These systems are in use in Minnesota, England, and Philadelphia. Images are deleted after 15 minutes if no action is taken.

AI in Gaming

Anthropic tested its Claude 3.7 Sonnet AI model on Pokémon Red. The model successfully battled three gym leaders. Games are being used for AI benchmarking purposes.

AI Model Developments

New AI Models and Updates

• Anthropic has released Claude 3.7 Sonnet, a hybrid AI reasoning model. This model allows users to control the duration the AI "thinks" about questions, integrating reasoning abilities into a single model. It also excels in math and coding. The model features a "scratchpad" to reveal its reasoning process, and users can refine prompts. Along with Claude 3.7 Sonnet, they are launching Claude Code, an agentic coding tool for active code collaboration. This tool enables code search, editing, testing, and GitHub integration.

• xAI launched Grok3, its flagship AI model, leading to a surge in downloads and usage of the Grok app.
• DeepSeek, a Chinese AI lab, has gained prominence with its chatbot, with its DeepSeek-V2 family of models showing impressive performance.

Integration of AI in Existing Platforms

• Google's Gemini may soon be integrated into iPhones as an AI assistant, with code in the iOS 18.4 developer beta suggesting its integration. Users will likely be able to use a basic version for free or access their Gemini Advanced account.
• Perplexity AI plans to launch Comet, a browser specifically for agentic search, which combines AI with real-time web searching. It offers links as citations, making it a research tool.

AI Technology and Research

Technical Advancements in AI Models

• The Transformer architecture is noted for processing all words in the input sequence in parallel, boosting training efficiency and improving the model’s ability to handle long sequences of text. The self-attention mechanism allows the model to weigh the importance of different parts of a sequence when processing each word.
• The origins of large language models can be traced back to the inception of AI research, influenced by both symbolic and stochastic approaches.

AI in Cognitive Science and Philosophy

• Artificial neural networks, including NLP architectures, are central to philosophical inquiry, particularly among philosophers of mind, language, and science, focusing on their suitability to model human cognition.
• Large language models like GPT-4 have demonstrated proficiency in language-based tasks, prompting debates about their linguistic and cognitive competence.

Other

Intel Xeon 6 Processors

• Intel is launching its new Intel Xeon 6 processors with performance-cores, offering better performance across data center workloads and up to two times higher performance in AI processing.

Controversial AI-Generated Content

• An AI-generated video depicting Donald Trump kissing Elon Musk's feet was displayed on TVs at the Department of Housing and Urban Development (HUD) headquarters.

Government & Politics

Unusual Incident at HUD Headquarters

The Department of Housing and Urban Development (HUD) headquarters in Washington, D.C., faced a disruptive incident. Televisions displayed an AI-generated video depicting President Donald Trump kissing Elon Musk's feet, accompanied by the phrase "LONG LIVE THE REAL KING." Staff had to manually shut off the TVs to halt the video. This event happened shortly after leaked documents revealed plans to cut 4,000 HUD employees under Elon Musk's DOGE project. Furthermore, HUD employees received an email from the Office of Personnel Management asking for a summary of their achievements, with agency leadership advising them to postpone their responses.

Artificial Intelligence

ChatGPT User Growth

ChatGPT has surpassed 400 million weekly users, doubling its user base since August. The AI chatbot's popularity is due to its advanced writing skills, STEM task proficiency, and updates like the integration of LLM GPT-4. OpenAI's COO Brad Lightcap noted that over two million business users use ChatGPT at work and teased the upcoming availability of GPT-4.5 and GPT-5.

AI Development

Google's AI Models and Pricing

Google's Veo 2 model, an AI video generator, is priced at $0.50 per second of footage, with clips capped at two minutes in up to 4K resolution. This pay-per-clip model differs from ChatGPT's subscription-based approach. Veo 2 is being tested for YouTube Shorts backgrounds and could potentially impact the stock footage industry, although current stock footage generally boasts higher quality.

Google provides both free and paid options for its Gemini AI bot. The Gemini Advanced tier, accessible through a $20/month Google One AI Premium plan, utilizes superior AI models compared to the free version. All users benefit from features like file uploads, image generation, real-time chat, and connectivity with other Google apps. However, Gemini Advanced users can work with larger files and generate images including people. While basic AI capabilities are consistent across versions, the advanced model provides better detail and comprehensiveness.