OpenAI

OpenAI has officially begun rolling out its Advanced Voice Mode (AVM) to a broader set of ChatGPT’s paying users, starting with those in the Plus and Teams tiers. This new feature enhances the platform’s natural speech capabilities, and it is expected to reach Enterprise and Edu customers next week. Continue Reading →

OpenAI released a new AI model called o1, which the company says is the first in a "new series of reasoning models for solving hard problems." Is this the rumored project Strawberry? It has many of Strawberry’s rumored features. Let's dig in. Continue Reading →

According to multiple sources (including The Wall Street Journal and Bloomberg), Apple, Nvidia, and Microsoft are in discussions to participate in a new funding round (led by Thrive Capital) that would value OpenAI at over $100 billion. The exact amounts that Apple, Nvidia, and Microsoft might invest have not yet been reported. Continue Reading →

OpenAI is getting ready to launch "Strawberry," its latest AI model (previously known as Q*). This new model is said to possess capabilities far beyond current AI, including solving novel math problems, developing marketing strategies, and even tackling complex word puzzles. (We'll have to wait and see if the new model knows how many times the letter "r" appears in "strawberry.") Continue Reading →

Condé Nast has signed a multi-year deal with OpenAI, allowing the AI company to use content from its publications, such as The New Yorker, Vogue, and Vanity Fair. This partnership aims to generate revenue and ensure proper attribution for the media company’s content, amidst challenges posed by changes in traditional search engines. Details of the deal are undisclosed. Continue Reading →

I read this study before bed last night: "Conversational AI Powered by Large Language Models Amplifies False Memories in Witness Interviews." It scared me more than I thought it would. The study explores how chatbots powered by LLMs can influence the formation of false memories for users; in other words, they can literally make users hallucinate. Continue Reading →

It started in 2022 with perplexity.ai, which (when it was working great) allowed you to do 10 minutes of Google searches in about 10 seconds. Google's answer? AI Overviews, which got everyone's knickers in a knot because there were no links, which meant no revenue for content creators or publishers. Now, there's OpenAI's newly announced SearchGPT, which will not only presents links, but also organizes and contextualizes information with summaries, insights, and detailed answers. To quote Yoda: "Begun, the search wars have." Continue Reading →

OpenAI is set to launch GPT-4o mini, a more compact version of its advanced AI model, later today. The model (which the company calls “the most capable and cost-efficient small model available today”) is designed for high efficiency and low cost, with planned future enhancements to support image, video, and audio integrations. Continue Reading →

How smart is any particular AI model? How would you set benchmarks? How would you rate them? AI models can't have an IQ; that's a test for human knowledge. (Plus, raw intelligence isn't really what you want to test anyway.) There are far more capabilities you'd want to include in your scoring system, such as safety, alignment with human values (whatever they are), etc. It's a monumental challenge – and OpenAI is taking it on. Continue Reading →

I forgot to mention a super fun moment from Cannes last week. Zombie brand Toys"R"Us and OpenAI flexed a bit as they showed the "first-ever brand film" (aka commercial) using Sora, OpenAI's new (yet unreleased) text-to-video tool. Continue Reading →