Use AI as Your Personal Assistant

Ready to embrace a new era of task delegation?

HubSpot’s highly anticipated AI Task Delegation Playbook is your key to supercharging your productivity and saving precious time.

Learn how to integrate AI technology into your processes, allowing you to optimize resource allocation and maximize output with precision and ease.

Download Now!

Today:

• Crusoe Announces $3.4 Billion AI Center

• Adobe Previews 3D, Animation AI

• Former Palantir CISO Joins OpenAI

• NYT Issues Cease Notice To Perplexity

• U.S. Mulls Expanding AI Chip Limits

• Trademark Clash: OpenAI vs Open AI

Crusoe Announces $3.4 Billion AI Center

Crusoe Energy Systems, Blue Owl Capital, and Primary Digital Infrastructure have formed a $3.4 billion joint venture to develop an AI data center at Lancium Clean Campus in Abilene, Texas. The 206 MW, 998,000 square-foot facility will support high-performance AI workloads with energy-efficient design, capable of handling 100,000 GPUs. Powered by renewable energy, the center aims to advance sustainable AI infrastructure. 

The project will create nearly 100 local jobs and contribute $1 billion to the economy over 20 years. The center is set to open in 2025, addressing the rising demand for AI computing power.

Adobe Previews 3D, Animation AI

Adobe has previewed experimental AI tools designed to streamline content creation. "Project Scenic" allows users to build 3D scenes and convert them into 2D images with more control. "Project Motion" simplifies animation, letting users create animated graphics and then enhance them using text descriptions. "Project Clean Machine" removes distractions from videos and photos, correcting visual inconsistencies like camera flashes or unwanted objects. 

These AI tools, revealed as part of Adobe’s MAX 2024 "Sneaks," aim to automate complex tasks while giving creators more control, though public access is not yet available.

Former Palantir CISO Joins OpenAI

Dane Stuckey, former CISO of Palantir, has joined OpenAI as its new Chief Information Security Officer (CISO). Stuckey brings over a decade of experience in security roles across commercial, government, and intelligence sectors. At OpenAI, he will focus on maintaining high standards for security and compliance, crucial for protecting millions of users and advancing AI development. 

OpenAI has recently intensified its security efforts, including collaboration with the U.S. Department of Defense and seeking experts in AI infrastructure security. Stuckey's expertise may help OpenAI's ambitions in government and defense-related AI projects.

NYT Issues Cease Notice To Perplexity

The New York Times has issued a "cease and desist" notice to AI startup Perplexity, accusing it of using its content without permission for generative AI purposes, which they claim violates copyright law. The NYT asked Perplexity to explain how it accesses the publisher's website despite previous assurances that it would stop using crawling technology. 

Perplexity denies scraping content for building models, stating it indexes web pages for factual citations. The NYT is also in a legal dispute with OpenAI over similar content usage claims, reflecting growing tensions between publishers and AI firms over content rights.

U.S. Mulls Expanding AI Chip Limits

The Biden administration is considering expanding restrictions on the sale of advanced AI chips from Nvidia and AMD to more countries, including possibly Saudi Arabia and the UAE. This move, aimed at safeguarding national security, would cap export licenses to Persian Gulf countries that are heavily investing in AI, such as Saudi Arabia's $40 billion AI fund and the UAE's potential involvement in OpenAI's funding. 

The U.S. previously banned these chips' sales to China and restricted exports to 40 other countries. These actions reflect concerns over global influence in the AI industry.

Trademark Clash: OpenAI vs Open AI

The legal battle between OpenAI, known for developing ChatGPT, and a lesser-known company called Open Artificial Intelligence Inc. (Open AI with a space) revolves around a trademark dispute. OpenAI (without a space), backed by Microsoft with significant investments, has filed a lawsuit against Open AI over trademark infringement. 

This conflict, though appearing trivial due to the similar names, reflects deeper tensions within Silicon Valley's competitive tech landscape. It could also influence future developments in the AI industry regarding branding rights and legal precedents in technology.

🧠RESEARCH

Diffusion for World Modeling: Visual Details Matter in Atari

DIAMOND (Diffusion As a Model Of eNvironment Dreams) is a reinforcement learning agent trained in a diffusion world model to improve visual detail in games like Atari and CounterStrike. It achieves a 46% higher score than human players in Atari 100k by simulating game frames with precise visual accuracy.

LOKI: A Comprehensive Synthetic Data Detection Benchmark using Large Multimodal Models

LOKI is a benchmark designed to evaluate the ability of large multimodal models (LMMs) to detect synthetic data across various modalities, including video, image, text, and audio. With 18,000 questions and multiple difficulty levels, LOKI tests LMMs’ judgment, perception, and reasoning abilities. The evaluation of 28 models reveals both potential and limitations in detecting AI-generated content.

MMIE: Massive Multimodal Interleaved Comprehension Benchmark for Large Vision-Language Models

MMIE presents a large-scale benchmark designed to evaluate Large Vision-Language Models (LVLMs) on interleaved multimodal comprehension and generation tasks. MMIE features 20,000 queries across 12 fields, such as mathematics and arts, with both multiple-choice and open-ended formats. It introduces a reliable evaluation metric to reduce bias and assesses eight LVLMs, revealing significant room for improvement.

Semantic Image Inversion and Editing using Rectified Stochastic Differential Equations

The paper introduces a novel method for semantic image inversion and editing using Rectified Flows (RFs) and stochastic differential equations. It tackles challenges in image inversion with Diffusion Models (DMs) and offers a more efficient alternative through RFs without the need for additional parameter training. The proposed method enables zero-shot inversion and editing, achieving state-of-the-art performance in tasks like stroke-to-image synthesis and semantic editing, with superior results confirmed by human evaluations.

Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Models

Omni-MATH introduces a benchmark designed to test large language models (LLMs) on Olympiad-level mathematics, offering a more challenging alternative to existing benchmarks like GSM8K. With 4,428 competition-level problems across 33 sub-domains, the benchmark highlights that even advanced LLMs like OpenAI o1-mini struggle, achieving around 60% accuracy, revealing gaps in high-level mathematical reasoning.

🛠️TOP TOOLS

PicLooks - Authentic profile pictures for your designs

Height.app - Autonomous project management tool

Quizzio - Personalized quizzes generated from your study material.

Clipdrop - Create stunning visuals in seconds

LLMWare - AI for Complex Enterprises

📲SOCIAL MEDIA

🗞️MORE NEWS

• Google Shopping has been rebuilt with AI, offering a personalized shopping experience. It provides AI-generated recommendations, virtual try-ons, price comparison, and deal-finding tools. Users can customize preferences and enjoy more efficient product searches and personalized deals.

• YouTube regularly tests new features with small groups of users to improve the platform. Recent experiments include AI-enhanced comment reply suggestions, a "Save" button for Shorts, and editing options for videos violating Community Guidelines.

• Amazon Ads introduced two new AI tools: AI creative studio and Audio generator, to simplify ad creation for advertisers. These tools help produce high-quality ads, including images, videos, and audio, enhancing customer engagement and campaign performance.

• Former OpenAI CTO Mira Murati has left the company and is reportedly discussing a new venture with OpenAI employees, though details remain undisclosed. Her departure adds to the internal challenges at OpenAI, where several researchers have requested team changes.

• OpenAI is restructuring its complex setup, aiming to separate its for-profit business from the nonprofit charity that oversees it. The charity could soon be worth $40 billion, and a new board may be created to govern the 501(c)(3) charity independently.

• Anthropic updated its Responsible Scaling Policy to address AI risks as models grow more powerful. New Capability Thresholds ensure additional safeguards for risky AI developments, such as bioweapons creation and autonomous AI research, aiming to prevent misuse.

• Amazon and Databricks have struck a five-year deal to use Amazon's Trainium AI chips, aiming to reduce costs for businesses building AI systems. This partnership expands on their existing collaboration, competing against Nvidia's dominant AI chips.