Sakhanda Wire
NVDA $155.46 -0.44% MSFT $506.24 -2.27% GOOGL $280.20 -1.24% META $618.87 -1.20% AMZN $257.89 +0.05%

Latest news

Older stories

TechCrunch

OpenAI opens Decisions API as fast decision models gain momentum

OpenAI has made its Decisions API available to all customers, bringing its decision-model offering into a market created around fast, low-cost systems that select among options and provide confidence estimates. The company entered the field after TypeSafe’s Jev model helped establish the category and contributed to the startup reaching a $7.6 billion valuation.

TechCrunch

Canva launches its first image generation model, Canva Craft

Canva has released Canva Craft, its first image generation model, developed by the company’s research team. The model is designed to compete with systems from Google, OpenAI, ByteDance and Meta, and is available to all users through the Dream Lab tab. Canva also plans to offer it through its API and support image generation in third-party chatbots such as ChatGPT and Gemini through Canva’s MCP.

The Verge

Discord improves Markdown previews and adds file and server tools

Discord’s latest update changes how uploaded Markdown (.md) files appear in previews. The company says they should now be rendered cleanly rather than displayed as a collection of symbols. The updated view supports tables and other formatting, which Discord says may make it easier to share README files and information related to AI projects.

MIT Tech Review

Building the foundation for scalable AI

A session at MIT Technology Review’s EmTech Future conference examines the conditions needed to develop scalable artificial intelligence in a highly regulated environment. It presents data, engineering, platforms, operating models and governance as connected elements that must function together rather than as separate projects.

TechCrunch

Microsoft AI outperformed generalist doctors in complex case simulation

Microsoft’s MAI-DxO diagnostic system reached an accuracy rate of 83% when tested on complex cases from the New England Journal of Medicine, according to a study published in Science. Primary care doctors and general internists, working individually with reference materials, achieved 36%. MAI-DxO coordinates several AI agents that question a simulated patient, request tests and discuss possible diagnoses. The comparison did not include specialists or physician teams, who typically handle cases of this difficulty.

The Decoder

Major US banks restrict AI agents using customer login credentials

AI agents are accessing bank accounts with credentials supplied by customers, despite policies that require account holders to log in themselves. Morgan Stanley, JPMorgan, Bank of America and Wells Fargo are among the major US banks restricting access by external agents. Some agents may also perform actions without receiving a direct instruction from the user.

The Verge

Zuckerberg bets Meta’s future on personal AI agent Muse

Mark Zuckerberg is presenting a future in which autonomous AI agents handle a growing share of everyday interactions. In that model, agents could shop and work on behalf of people, while also communicating with other people’s agents instead of their human owners dealing directly with one another.

The Verge

Microsoft claims AI diagnostic tool outperforms doctors

Microsoft says it has developed an artificial intelligence tool that can identify medical conditions up to four times more accurately than human doctors. Mustafa Suleyman, the company’s AI chief, described the development to The Verge as evidence that AI’s long-promised impact on healthcare is now emerging.

The Decoder

Google introduces isolated microVMs for Gemini agents

Google has announced Agent Compute Engine, a runtime designed to give every AI agent its own isolated virtual machine. The service already runs Gemini and is expected to become available for preview in the coming months. According to Google, each agent operates in a microVM protected by the company’s Titanium chip.

MarkTechPost

Liquid AI releases RIPL for low-latency LFM inference

Liquid AI has introduced RIPL, or Real-time Inference Pipeline, an inference engine designed to run Liquid Foundation Models in live request paths. The company says the system can return complete generated responses in under 50 milliseconds and maintain that latency under continuous peak traffic. RIPL supports LFM2.5 models ranging from 230M to 2.6B parameters on enterprise deployments using NVIDIA H100, H200, B200 and RTX PRO 6000 GPUs.

The Verge

Modder demonstrates apparent Switch 2 jailbreak using an older game

The Nintendo Switch 2 has apparently been jailbroken, with modder Sammy Hajhamid, known online as “pepsipu,” demonstrating a customized version of Mario Kart World. The build adds playable characters not included in the standard game, including Cinnamoroll, the red Angry Bird, Peter Griffin and Steve from Minecraft.

MarkTechPost

Gradium open-sources 31M-parameter Phonon speech model

Gradium has released Phonon, a family of compact text-to-speech models designed to generate audio locally on phones, computers, Raspberry Pi 5 boards and browsers. The English Phonon 31m has 31 million parameters, produces 24 kHz mono audio and is distributed under Apache 2.0. Its quantized GGUF file is 64.5 MB, while the Phonon 40m file is 74.6 MB.