News

The Week in AI: Announcements from OpenAI, Microsoft, Intel, Google, NVIDIA, Stability AI, More

• By John K. Waters
• 02/26/2024

This edition of our weekly roundup of AI product-and-services news includes OpenAI's GPT new rating feature in the OpenAI Store, Microsoft's decision to go with Intel for its AI chips, Google's apology for its Gemini Stumble, NVIDIA's release its new NVIDIA App Beta, Stability AI's preview of Stable Diffusion 3, Samsung's announcement of new AI features on more Galaxy devices, Figure AI's lineup of big tech investors, and the inclusion of three AI startups in Disney's 10th Accelerator Program.

Third-Party GPT Users Get to Rate them in the OpenAI Store
OpenAI announced an update to its GPT Store that allows users of third-party GPTs to rate them using a dropdown menu that lists a new option : "Send feedback." This Consensus option opens a new interstitial pop-up interactive element that allows users to leave a one- to five-star rating and a text box to send an "optional private email" to that GPT’s creator. There's also a checkbox option where the rater can leave his or her email address. OpenAI made the announcement on the X social media platform; the post includes an animated GIF demonstrating this new option.

AI-Enabled Robots Maker Figure AI Attracts Big Backers
A startup that's using AI in its development of human-like robots is attracting a list of big-name investors, including Nvidia, Jeff Bezos, Microsoft, and OpenAI. The company, Figure AI, is raising about $675 million in a funding round that carries a pre-money valuation of roughly $2 billion, Bloomberg reported last week. Microsoft is reportedly contributing $95 million, and NVIDA, along with an Amazon-affiliated fund, are investing $50 million each, Bloomberg reported. Bezos has committed to $100 million via Explore Investments, LLC. Figure AI aims to position its AI-powered, human-shaped Figure 01 robot in warehouses.

Microsoft Picks Intel for AI Chips
Intel, the world's largest maker of central processing units and semiconductors, will soon be making high-end chips for Microsoft's artificial intelligence efforts, the companies announced last week. Chief executive Pat Gelsinger said at a recent company event that his company is set to "rebuild Western manufacturing at scale." The comment refers to Intel's reorganization into two units: Intel Foundry, which manufactures semiconductors, and Intel Products, which is responsible for their design.

Google Says Sorry for Gemini Image-Generation Stumble
Google apologized on Friday for the flawed rollout of its Gemini AI image-generator, acknowledging that, in some cases, the tool would "overcompensate" in its effort to find a diverse range of people even when such a range didn’t make sense. The apology, posted on X, reads: "We’re working to improve these kinds of depictions immediately. Gemini’s AI image generation does generate a wide range of people. And that’s generally a good thing because people around the world use it. But it’s missing the mark here." The company announced that it was suspending Gemini AI's ability to generate pictures of people until the flaw is fixed.

"It’s clear that this feature missed the mark," said Prabhakar Raghavan, senior VP in Google's Knowledge and Information group, in a blog post "Some of the images generated are inaccurate or even offensive. We’re grateful for users’ feedback and are sorry the feature didn’t work well."

Google recently added the new image-generating feature to its Gemini chatbot, formerly known as Bard, about three weeks ago. It was built on an earlier Google research experiment called Imagen 2.

Stability AI Previews Stable Diffusion 3
Stability AI announced an early preview release of its Stable Diffusion 3 last week. The company described the release as "our most capable text-to-image model with greatly improved performance in multi-subject prompts, image quality, and spelling abilities." The model is not yet broadly available, but the company is opening the waitlist for this early preview. This preview phase was launched so the company can gather "crucial insights" to improve the model's performance and safety ahead of an open release.

Stable Diffusion 3 combines a diffusion transformer architecture and flow matching. Its suite of models currently ranges from 800M to 8B parameters. This approach aims to align with Stability AI's core values and "democratize access, providing users with a variety of options for scalability and quality to best meet their creative needs." The company says it plans to publish a detailed technical report, soon. Anyone interested in joining the early preview can find more information here. 

NVIDIA Releases New App Beta
NVIDIA announced the release of a public beta of the new NVIDIA App, which the company calls "the essential companion for gamers and creators with NVIDIA GPUs in their PCs and laptops." The NVIDIA app beta is a first step on the NVIDIA's journey to modernize and unify the NVIDIA Control, PanelGeForce Experience, and RTX Experience apps, the company said.  
Featuring a unified GPU control center, NVIDIA app allows fine-tuning of game and driver settings from a single place, while introducing a redesigned in-game overlay for convenient access to powerful gameplay recording tools, performance monitoring overlays, and game enhancing filters, including innovative new AI-powered filters for GeForce RTX users.

This initial beta release incorporates many of the top features from our existing apps, optimizes the user experience, includes an optional login to redeem bundles and rewards, and introduces new RTX capabilities to elevate your gaming and creative experiences.

Samsung Puts More AI in Galaxy Devices
Samsung Electronics  announced the availability of new AI features on more Galaxy devices, last week. Available through a new One UI 6.1, the update was engineered to further the "democratization" of mobile AI. The update will be available across the Galaxy S23 series, S23 FE, Z Fold5, Z Flip5, and Tab S9 series, the company said, and the roll out will start at the end of March. Aligning with the recently launched Galaxy S24 series, this update "elevates the standard of users' mobile AI experience through a hybrid approach that combines on-device and cloud-based AI," the company said.

"Our goal with Galaxy AI is not only to pioneer a new era of mobile AI but also to empower users by making AI more accessible," said TM Roh, President and Head of Mobile eXperience Business at Samsung Electronics, in a statement. "This is only the beginning of Galaxy AI, as we plan to bring the experience to over 100 million Galaxy users within 2024 and continue to innovate ways to harness the unlimited possibilities of mobile AI."

Disney's Accelerator Program Includes Three AI Startups
The Walt Disney Company has unveiled the lineup for its 10th Disney Accelerator program, a business development initiative established "to aid and accelerate the growth of companies deemed by Disney to be doing innovative work." The program grants selected companies investment capital, access to co-working space at Disney’s creative campus, and mentor support and guidance from top Disney executives.
The list of five companies selected to participate in this year’s Accelerator program include three specializing in AI tools and technology, one in virtual reality, and another in autonomous vehicles. This year’s participants are:

• Audioshake: An audio technology company that uses AI to separate layers of recorded sound to make audio interactive, editable, and customizable.
• Elevenlabs: A voice AI research and deployment company that creates realistic, versatile, and contextually aware AI audio.
• Nuro: An autonomous vehicle company that builds custom, electric, zero-occupant vehicles for the delivery of goods.
• PrometheanAI: A company that provides a suite of tools for virtual world creation and digital asset management using natural language prompts.
• Statuspro: An immersive entertainment company that leverages virtual and augmented reality to create first-person sports gaming experiences.

Mistral Joins Amazon Bedrock's Roster of AI Foundation Models
Startup Mistral is extending access to its two open source large language models to developers using Amazon's prodigious cloud platform. The models, Mistral 7B and Mixtral 8X7B, will be available on the Amazon Bedrock developer platform "soon," according to an AWS blog post last Friday, though no specific release date was given. Both models are available under the Apache 2.0 license, and are designed to be lightweight, fast and customizable. (See our coverage here.)