đ OpenAI 100x's Itself, Home Assistance Robots Are Here, Elon Musk's Colossus, AI & Crispr, Early Cancer Detection
GPT-Next 100x Improvement Over GPT-4, NEO Beta Robot Happily Does Your Laundry & Dishes, Elon's 100,000 Nvidia H100 GPUs, New Breakthroughs in Bio-AI
Welcome to this weekâs edition of AImpulse, a five point summary of the most significant advancements in the world of Artificial Intelligence.
Hereâs the pulse on this weekâs top stories:
Whatâs Happening: OpenAI Japanâs CEO just stated (article is in Japanese) that âGPT-Nextâ, OpenAIâs upcoming AI model, is expected to be 100 times more powerful than GPT-4, while also confirming a more advanced model named Orion is slated for 2025.
The details:
GPT-Next is expected to achieve a 100x performance increase without wasting significantly more computing resources.
The improvement comes from better architecture and learning efficiency, not just raw computing power.
GPT-Next will use a smaller version of âStrawberryâ, an advanced AI system OpenAI is currently developing.
The slide mentions GPT Next 2024, which while unconfirmed, hints that the model may be released by the end of 2024.
Why it matters: OpenAI is still the undisputed leader in AI model usage, and if translations are accurate, GPT-Next will be the next huge leap that developers have been waiting for. Most impressively, it makes that leap without using much more compute, a major bonus for apps that require speed + efficiency.
Whatâs Happening: 1X Technologies just unveiled the most realistic AI-powered humanoid yet: NEO Beta, an advanced robot designed for everyday home assistance with human-like capabilities in movement, interaction, and task performance.
The details:
NEO Beta stands 5 feet 5 inches tall, weighs 66 pounds, and can walk at 2.5 miles per hour and run at 7.5 miles per hour.
The robot has a carry capacity of 44 pounds and can operate for 2 to 4 hours on a single charge.
NEO Beta uses embodied artificial intelligence to understand its environment and learn from past interactions.
Itâs designed to perform a wide range of tasks, from household chores to providing companionship and support for individuals with mobility challenges, and it can be remotely operated by a human if necessary.
Why it matters: The race to an affordable robot butler is officially underway and 1X Technologies just boosted itself into the same conversation as Chinaâs AGIBOT fleet and Teslaâs Optimus. NEO is so realistic and smooth that thousands of people on X (Twitter) were debating if it was a person in a suit.
Whatâs Happening: Elon Muskâs xAI just launched âColossusâ, the worldâs most powerful AI cluster powered by a whopping 100,000 Nvidia H100 GPUs, which was built in just 122 days and is planned to double in size soon.
The details:
Colossus consists of 100k liquid-cooled Nvidia H100 GPUs connected on a single network fabric and is considered the âmost powerfulâ in the world, according to Musk.
The system was built and launched in just 122 days, an incredibly quick accomplishment that was brought online over a holiday weekend in the United States.
xAI plans to expand Colossus to 200,000 GPUs (50,000 H200s) in the coming months.
In a recent podcast, Elon Musk (founder of xAI) said Grok 2 was trained on only around 15,000 GPUs.
Why it matters: xAIâs Grok 2 recently caught up to OpenAIâs GPT-4 in record time, and was trained on only around 15,000 GPUs. With now more than six times that amount in production, the xAI team and future versions of Grok are going to put a significant amount of pressure on OpenAI, Google, and others to deliver.
Whatâs Happening: Researchers at Profluent Bio used AI to successfully create OpenCRISPR-1, a novel gene-editing tool that matches or surpasses the performance and safety of current CRISPR systems.
The details:
Researchers used Large Language Models (LLMs) trained on over a million CRISPR operons to understand and generate novel protein structures.
The LLMs designed a new gene-editing tool called OpenCRISPR-1, capable of various tasks including base editing.
In tests, OpenCRISPR-1 showed a similar efficiency in cutting DNA to the current models while making fewer mistakes.
Researchers made OpenCRISPR-1âs sequence publicly available to encourage broader use in research.
Why it matters: OpenCRISPR-1 helps to treat genetic diseases in the safest, fastest, and most personalized way possible. While still years away from practical applications, AI tools like this opens doors for custom gene-editors tailored to individual DNA, potentially accelerating cures for previously untreatable conditions.
Whatâs Happening: Researchers recently developed an AI tool called AINU that can differentiate cancer cells from normal cells and detect early stages of viral infection, by analyzing high-resolution images of cell nuclei.
The details:
AINU uses a convolutional neural network to analyze images captured by STORM microscopy, which offers nanoscale resolution.
The AI can detect structural changes in cells as small as 20 nanometers, 5,000 times smaller than a human hair's width.
AINU also detected viral infections (herpes simplex virus type-1) just one hour after infection by observing subtle changes in DNA packing.
The tool can accurately identify stem cells too, which could accelerate stem cell research without relying on animal testing.
Why it matters: Yesterday, researchers revealed an AI tool to help with early dementia detection, and now AI is detecting cancer cells at a nanoscale level. Clinical applications may be years away, but AI healthcare breakthroughs like AINU are only accelerating â and will dramatically revolutionize scientific research in the coming years.