Dark Mode Light Mode
US Flu Vaccine Composition Uncertain Amidst FDA Meeting Cancellation
OpenAI Unveils GPT-4.5: A More Natural and Knowledgeable AI Model

OpenAI Unveils GPT-4.5: A More Natural and Knowledgeable AI Model

OpenAI Unveils GPT-4.5: A More Natural and Knowledgeable AI Model OpenAI Unveils GPT-4.5: A More Natural and Knowledgeable AI Model

OpenAI has introduced GPT-4.5, its latest and most advanced AI model to date. While not a reasoning model like the O-series, GPT-4.5 excels in natural conversation, boasting a higher emotional quotient and improved problem-solving skills.

This new model offers access to up-to-date web information, facilitates file and multimedia uploads, and supports canvas platforms for coding tasks. However, it currently lacks voice mode, video comprehension, and screen-sharing capabilities.

GPT-4.5 is initially limited to a research preview phase, restricting access even for ChatGPT Plus subscribers. OpenAI CEO Sam Altman attributed this limited availability to infrastructure challenges, specifically a shortage of GPUs, which the company plans to address soon. Altman stated, “It is a giant, expensive model… we’ve been growing a lot and are out of GPUs. We will add tens of thousands of GPUs next week and roll it out to the Plus tier then.”

See also  10 Landmark AI Advancements of 2024

Unlike models employing chain-of-thought reasoning, GPT-4.5 utilizes unsupervised learning, scaling both training data and computational resources. This approach leads to reduced hallucination (a major concern with generative AI chatbots), a broader knowledge base, and deeper understanding.

Altman tweeted his excitement about the new model, highlighting its conversational abilities while acknowledging the infrastructure challenges related to its size and cost.

GPT-4.5 is ready! Good news: it is the first model that feels like talking to a thoughtful person to me… Bad news: it is a giant, expensive model…

— Sam Altman (@sama) February 27, 2025

Benchmarks released by OpenAI indicate that GPT-4.5 surpasses GPT-4o, o3-mini, and o1 in accuracy, while simultaneously exhibiting a significantly lower hallucination rate. OpenAI explains, “For GPT‑4.5, we developed new, scalable techniques that enable training larger and more powerful models with data derived from smaller models.” This allows for more natural handling of creative queries, professional questions, and nuanced conversations.

See also  Lossless Scaling 2.12 Beta: Enhanced Frame Generation Surpasses DLSS 3 and FSR 3

OpenAI asserts that GPT-4.5 excels in creativity, aesthetic intelligence, writing, and design-related tasks. Access is currently rolling out to ChatGPT Pro subscribers across web, mobile, and desktop platforms. ChatGPT Plus and Team tiers will gain access next week, followed by Enterprise and Education bundles.

Add a comment Add a comment

Leave a Reply

Your email address will not be published. Required fields are marked *