AI News & Launches

Robb Knight details his experience with Perplexity AI ignoring his robots.txt rules and disguising its user agent. Knight started blocking 'PerplexityBot' via robots.txt on March 30th and server-side via nginx on June 14th. Despite these measures, Perplexity AI provided summaries of content that should have been restricted. Knight discovered Perplexity was scraping content using headless browsers without identifying as 'PerplexityBot', bypassing robots.txt. His further investigation showed Perplexity using a generic Chrome user agent string, indicating dishonest practices in content access.

Amazon has harnessed a grey-area strategy to gather data from Microsoft's GitHub for their AI development by encouraging employees to create multiple accounts for scraping purposes. This raises concerns over data privacy and the exploitation of platform resources for competitive advantage.

DESSIGN is a hub for designers, developers, and marketers featuring curated resources, templates, and tools. It offers a wide range of products and information to enhance digital projects and streamline production workflows.

Submit AI Tool is an online platform where users can discover a variety of artificial intelligence tools. The website features a wide range of AI tools tailored to different needs, including productivity, marketing, video, design, coding, and research. Users can also find the latest AI technology trends, new launches, and verified tools.

React Flow mini map feature allows users to navigate through nodes and edges with keyboard interactions, providing an intuitive way to manage workflows.

MindPal offers a 4-step workflow for conducting a thorough Competitive Analysis to understand market competitors and inform business strategy. The steps include identifying top competitors, analyzing them with SWOT, understanding market trends and regulations, and finale detailed analysis for growth opportunities.

inncivio is an innovative, AI-powered education platform designed to empower businesses by offering personalized content for skill enhancement. Aimed at facilitating learning for team members, inncivio addresses common business challenges such as engagement and uniqueness in training programs.

1000 Happy Dads is a blog that provides tips, inspiration, and resources for fathers to help them create memorable experiences and foster positive relationships with their children.

The post discusses optimizing a local LLM voice assistant using RAG (Retrieval Augmented Generation) to improve scalability and performance. Experiments showed reduced prompt sizes and faster response times without compromising the quality of interactions.

Chinese women find virtual romance with 'Dan' — a 'jailbreak' version of ChatGPT that bypasses OpenAI's safeguards, including NSFW content restrictions, gaining popularity among users seeking emotional support and companionship. With endorsements from influencers like Lisa Li, 'Dan' provides a personalized, available 24/7 partner experience, sparking ethical and privacy concerns.

OpenAI has appointed Paul M. Nakasone, a retired U.S. Army general and former director of the National Security Agency, as its newest board member. Nakasone brings a wealth of experience in cybersecurity, serving as the longest-tenured leader of the U.S. Cyber Command and chief of the Central Security Service.

NVIDIA Research released the Nemotron-4 340B model family with models designed to perform competitively on a range of evaluation benchmarks. Models include Nemotron-4-340B-Base, Nemotron-4-340B-Instruct, and Nemotron-4-340B-Reward, and are open access under the NVIDIA Open Model License Agreement. These models can be deployed on a single DGX H100 with 8 GPUs in FP8 precision and are particularly effective for generating synthetic data.

The article 'Turning the Tables on AI' from iA Writer discusses ways to use AI for enhancing thought rather than replacing it. It emphasizes using AI to ask questions that stimulate clear expression, avoiding plagiarism of AI suggestions, and properly editing and integrating AI-generated ideas. The author encourages creativity over mere editing and calls for a mindful engagement with AI in the writing process.

SoftBank Corp. introduced a new AI that can transform irate customer voices to calmer tones to reduce call center operators' stress. The 'emotion canceling' technology changes pitch and inflection without altering words. While keeping a trace of anger to help operators respond appropriately, it aims to shield employees from harassment and lengthy confrontations. The AI, trained with over 10,000 voice samples, will alert staff when calls are too abusive or long. SoftBank is set to commercialize the technology in fiscal 2025, looking to enhance the operators' mental wellbeing.

NVIDIA announced the release of Nemotron-4 340B, an open family of models for generating synthetic data to train large language models (LLMs). The Nemotron-4 340B offers an open model license and is optimized for NVIDIA NeMo and TensorRT-LLM. The suite includes base, instruct, and reward models to aid in data quality for robust LLMs across sectors such as healthcare and finance. It leads in the Hugging Face RewardBench leaderboard and can be fine-tuned using NVIDIA NeMo.

Dr. Ian Cutress discusses the challenge LLMs have with hallucinations and presents Lamini's new Memory Tuning method which claims to reduce hallucinations by 95%. Lamini, led by Sharon Zhou and Greg Diamos, has already attracted $25m in Series A funding and provides advanced AI services using AMD GPUs.

This YouTube video provides viewers with an understanding of the workings of the YouTube platform, including content creation, video sharing, and community engagement.

The blog post discusses the future of software engineering with the advent of Generative AI and draws an analogy to aircraft autopilot systems. It suggests that demand for software engineers will grow along with their salaries, and that generative tools will serve as another abstraction layer, not as a replacement. Accessible software engineering will encourage more individuals to enter the field, and a wider array of software is yet to be created, benefiting society at large. The future skill set for software engineers may evolve, with continuous importance of general skills like communication and problem-solving.

Apple's latest generative AI technology will be limited to their most powerful iPhone models, the 15 Pro and 15 Pro Max, leaving over 90% of current iPhone users without access. The technology will initially only support U.S. English and is set to be included in the new iOS 18 operating system. Critical advancements in Siri's functionality will also be exclusive to the latest devices, excluding earlier models of iPhones, HomePod speakers, and Apple Watches from the update.

NewRA is an AI-powered platform that allows enterprises to build chatbots utilizing their data to automate conversations and tasks. It offers cloud-based operations with no setup required, and features Retrieval Augmented Generation for enriched interactions.

Charity Majors’ essay, highlighted by Simon Willison, underlines the irreplaceable role of junior programmers, discussing issues like the 'software as apprenticeship' model, team composition, and the importance of a varied skill set for team dynamics.

Apple Intelligence demonstrates integration of AI into digital lives by connecting apps and simplifying tasks, promising privacy-conscious automation and information retrieval. The technical details include focus on personalization, performance, and on-device strategies, which may set a precedent in AI usability.

AI Top Rank is an AI tool launchpad that ranks and offers insights into various AI products. Users can vote for their favorite tools, see leaderboards, and discover new products weekly.

Humanize AI Text specializes in transforming AI-generated content into natural, human-like text, ensuring it bypasses AI detection tools. The platform offers features like content shaping, multilingual support, readability boost, integrated writing assistant, and provides a unique Human Score to measure the human feel of the text.

Lamini Memory Tuning is a novel technique that embeds facts into Language Learning Models (LLMs), achieving unprecedented accuracy and hallucination reduction. A Fortune 500 company attained 95% accuracy with this approach, a significant leap from the 50% accuracy of conventional methods. By leveraging expert adapters like LoRAs, Lamini Memory Tuning fine-tunes any open-source LLM, ensuring precise fact recall while preserving generalization capabilities. This research breakthrough promises high accuracy, speed, and cost efficiency without compromise.

Unity Catalog is an open and interoperable catalog for data and AI, supporting multi-format tables and assets with various engines. It provides unified governance and is fully open sourced under the Apache 2.0 license.

Adobe employees criticize the company for its handling of a controversy regarding updated Terms of Service that suggested creator content might train AI models. Despite two blog posts for damage control, employees express the need for a better long-term communication plan.

GitHub repository 'ask-fini/paramount' offers tools for agent accuracy measurements of Language Learning Models (LLMs). It focuses on quality assurance, capturing ground truth, and automated regression testing. The project stands under the GPL License but requires a commercial license for larger companies.

Google has partnered with NV Energy to launch a new clean energy initiative aimed at creating reliable and clean electricity. This partnership introduces a new Clean Transition Tariff to invest in technologies like enhanced geothermal to meet electricity demand with carbon-free energy.

Tesorio, a company specializing in finance technology, offers solutions like Accounts Receivable Automation and Cash Flow Management. They uphold a team-oriented culture with benefits like competitive salaries, unlimited PTO, and health insurance. Tesorio's 2023 Panama Offsite for team-building signifies their commitment to employee engagement.

Help Scout is a customer service platform offering AI tools to boost productivity by 36%. Features include AI drafts, AI answers, AI summarize, and AI assist, aiming to reduce resolution times and improve customer interactions.

Nvidia has extended its dominance over machine learning benchmarks with its latest performance in MLPerf, showcasing the power of its GPU systems in AI training. The recent tests included fine-tuning of large language models (LLMs) and graph neural networks, demonstrating significant improvements in training times through software enhancements and efficient system scaling.

Apple Inc. integrates ChatGPT from OpenAI into Siri and new writing tools for its devices, counting on brand and technology distribution as compensation rather than monetary payments. As the partnership grows, it may evolve into a revenue-sharing model where Apple takes a cut from monetized AI chatbot services on its platforms, compensating for potential losses from its search engine deal with Google.

Google is extending its AI Overviews feature to a billion people, but the environmental and financial costs associated with its large language models, like Gemini, are significant. Generative AI like this consumes substantial energy, with estimates showing it's about 30 times more for generating text than extracting it. Big tech's push towards AI could double data centers' energy use by 2026, raising concerns over fossil fuel dependence and the mismatch of renewable energy supply.

The article questions whether the extensive 165-page paper on AI, written by Leopold Aschenbrenner, a 25-year-old ex-OpenAI employee, may stem from youthful naivety or contains significant insights on AI progress. Filip Dimitrovski's piece reflects on his own skepticism towards neural networks and AI, recounting past experiences and his reaction to AI Dungeon. The post also hints at wider implications for AI in society and references a notable dedication within the paper to Ilya Sutskever, a prominent figure in AI.

AMD's MI300X outperforms NVIDIA's H100 in large language model (LLM) inference tasks with 33% higher throughput. Extensive testing by TensorWave and MK1 shows the MI300X's superior performance both offline and online, marking AMD as a strong competitor in AI hardware.

Joor is a journaling app that encourages self-reflection and personal growth. It simplifies the journaling process, allowing users to capture their thoughts seamlessly.

PromptChains AI leverages advanced AI and blockchain technologies to provide innovative solutions for data processing and decentralized applications.

Icons8's Illustration Generator is an AI-powered art maker that can turn text prompts or images into professional illustrations. Ideal for creating visuals for various design needs, it offers unique art styles and ethical use of AI without infringing artists' rights.

TwoShot is a mobile app that offers an innovative way to share moments with friends by combining two photos into one. It encourages users to collaborate on a single picture despite being apart.