OpenAI CEO Sam Altman is seeking to raise $5 trillion to $7 trillion for AI chip manufacturing to address the scarcity of GPUs for language models like ChatGPT and Microsoft Copilot. Altman is pitching a partnership between OpenAI, investors, chip makers, and power providers to build chip foundries, with OpenAI as a significant customer.
Hector Xu, a former MIT student, founded Rotor Technologies to make helicopter flight safer by retrofitting existing helicopters with autonomous technology. Rotor's autonomous aircraft can fly faster and longer than drones and are already conducting demo flights.
Recommender systems generate significant revenue, with Amazon and Netflix relying heavily on product recommendations. This article explores the use of controlled vocabularies and LLMs to improve similarity models in recommender systems, finding that a controlled vocabulary enhances outcomes and building a genre list using an LLM is easy but creating a detailed taxonomy is challenging.
The article "Principal Component Analysis (PCA) from Scratch Using the Classical Technique with C#" in Microsoft Visual Studio Magazine explains how PCA can reduce the number of columns in a dataset and its applications in machine learning algorithms. It also discusses the difficulty of computing eigenvalues and eigenvectors and provides a demo using a subset of the Iris dataset.
Learn how to create a custom AI using OpenAI's Assistants and Fine-tuning APIs in this step-by-step guide. Build an AI assistant with knowledge retrieval capabilities, like a YouTube comment responder, using the Assistants API.
Google has renamed its AI assistant Bard to Gemini and launched its most capable AI model, Ultra 1.0, as part of a $20/month subscription feature. The nomenclature and accessing the new model can be confusing, but it offers different AI "engines" for enhanced performance.
Automate mortgage document fraud detection using ML models and business-defined rules with Amazon Fraud Detector, a fully managed fraud detection service. Upload historical data, train the model, review performance, and deploy the API to make predictions for improved fraud detection and underwriting accuracy.
This article explores three key encoding techniques for machine learning: label encoding, one-hot encoding, and target encoding. It provides a beginner-friendly guide with pros, cons, and Python code examples to help data scientists understand and implement these techniques effectively.
Meta, the parent company of Facebook, plans to label AI-generated images from other companies like OpenAI and Google to enhance transparency on platforms like Facebook and Instagram. The move comes during a contentious US election year and is part of a larger effort to establish standards for labeling content created using generative AI models.
The article discusses the practical lessons learned from upgrading the Bed-Reader bioinformatics library to read DNA data directly from the cloud. The author provides nine rules for adding cloud-file support to programs, including using the object_store crate and creating a new crate called cloud-file.
Accenture has developed an AI-based solution using AWS services to automate the labor-intensive process of creating Common Technical Documents (CTDs) for pharmaceutical companies. This solution reduces authoring time by 40-45% and accelerates new product approvals, benefiting patients.
The MIT-Pillar AI Collective announces six fellows for spring 2024, supporting graduate students conducting research in AI, machine learning, and data science to commercialize their innovations. Fellows include Yasmeen AlFaraj, working on sustainable plastics, and Ruben Castro Ornelas, designing multipurpose robots with AI control.
The pharmaceutical industry generated $550 billion in US revenue in 2021, with a projected cost of $384 billion for pharmacovigilance activities by 2022. To address the challenges of monitoring adverse events, a machine learning-driven solution using Amazon SageMaker and Hugging Face's BioBERT model is developed, providing automated detection from various data sources.
This article highlights Mr_Vudoo, a digital renaissance man and featured artist in the NVIDIA Studio, who recently acquired a GeForce RTX 4080 SUPER graphics card to enhance his content creation. Additionally, Topaz Labs has adopted NVIDIA TensorRT acceleration for its AI-powered Remove Tool, delivering faster processing compared to other devices.
Doctors have lower accuracy in diagnosing skin diseases on darker skin, with dermatologists accurately characterizing only 34% of images compared to 38% for lighter skin. Artificial intelligence algorithms can improve accuracy, but the disparity suggests a need for changes in dermatology education and training.