Artificial intelligence’s impressive performance comes at the cost of significant energy consumption, with complex tasks demanding even more power. In response, researchers Víctor López-Pastor and Florian Marquardt from the Max Planck Institute for the Science of Light in Erlangen, Germany, have introduced an innovative approach to training AI that offers remarkable energy efficiency. Their method diverges from the conventional use of digital artificial neural networks and relies on physical processes instead. The results of their work are published in the journal Physical Review X.

Training models like GPT-3, the foundation of eloquent chatbots like ChatGPT, demands an astounding amount of energy. While the exact energy consumption for GPT-3 remains undisclosed by Open AI, estimates suggest it could equate to the annual consumption of 200 German households. Despite these energy expenditures, such AI models have yet to grasp the underlying meaning of phrases they analyze.

Singapore-based startup SpaceChain is launching a groundbreaking service that combines artificial intelligence with Earth observation data to provide valuable insights. The service, known as I-Sat and developed by SpaceChain’s U.S. subsidiary, SC Solutions, aims to streamline the process of extracting meaningful information from Earth imagery.

Ziheng Xiang, SpaceChain’s Director of Technology, explained that while AI has simplified the task of answering questions, most AI solutions lack real-time analysis and do not integrate seamlessly with Earth intelligence. I-Sat, akin to OpenAI’s ChatGPT, employs natural language processing technology to generate answers but goes a step further by incorporating real-time data analytics to support those answers.

In a recent report, Forrester delved into the future implications of generative artificial intelligence and how this technology is set to predominantly affect white-collar jobs.

The report defined generative AI as a collection of technologies and techniques that utilize extensive datasets, including large language models, to generate fresh content across various mediums such as text, video, images, audio, and code. This content generation can be initiated through natural language prompts or unconventional inputs.

Hiber, in collaboration with Datatonic and Google Cloud, announced a significant upgrade to its 3D world creation tools, incorporating generative AI to simplify the process of building virtual environments. This development was showcased at the recent Google Cloud Next event in San Francisco, marking a major milestone for the Swedish-based company.

Hiber’s mission is to democratize the 3D web, making it more accessible to creators and users alike. Through their existing tools, the Hiber3D Development Kit, and a consumer-friendly no-code tool, they have already facilitated the creation of over five million 3D worlds on the HiberWorld user-generated content hub.

Amazon has unveiled a new AI tool designed to assist sellers in creating product listings more efficiently. This feature utilizes a large language model (LLM) trained on extensive datasets to streamline the process of describing products. Amazon characterizes this tool as a way to simplify the often labor-intensive task of crafting titles, bullet points, and product descriptions into a single step.

Amazon’s Generative Listing Content tool requires sellers to provide a brief product description, typically just a few words or sentences. From there, the tool generates high-quality content, including a product title, description, and bullet points, which sellers can review and edit if necessary. During the testing phase, many sellers found the generated content suitable for use without further revisions.

Recent research conducted by Stanford University reveals that large language models (LLMs) like OpenAI’s GPT-3, 3.5, and 4 can be effectively harnessed to simulate human behavior convincingly. This innovative study, titled “Generative Agents: Interactive Simulacra of Human Behavior,” explores the potential of generative models in creating AI agents that mimic human behavior with dynamic realism.

Advancing Realistic AI Behavior

The study explores an AI agent architecture that remembers interactions, reflects on received information, and plans both short- and long-term goals based on an ever-expanding memory stream. These AI agents can simulate a wide range of human behaviors, from mundane tasks to complex decision-making processes. When these agents interact with one another, they can emulate intricate social behaviors, providing insights into population dynamics and societal interactions.