AI News

DeepMind Al's latest breakthrough, AlphaGeometry2, can outperform IO Math Olympiate geometry gold medalists, suggesting that AI-assisted mathematical reasoning is possible.AlphaGeometry2, a newer branch of DeepMind’s AlphaGeometry, surpassed other competitors after it achieved to solve 84% of all geometry problems presented within the last 25 years at the International Math Olympiad. DeepMind says that seeing a system solve geometry problems gives confidence that it will be able to reason through other tasks, which is very helpful in engineering and other theoretical sciences.

France’s public investment bank, Bpifrance, has announced plans to invest up to €10 billion ($10.3 billion) in the country’s AI ecosystem, reinforcing its ambitions to become a global AI powerhouse. The announcement comes just days before the Artificial Intelligence Action Summit in Paris, where world leaders and tech executives will discuss the future of AI.

Intel is halting Falcon Shores, its next-gen AI GPU, shifting focus to Jaguar Shores for a more strategic AI data center solution. The company admits to falling short in AI chip sales, facing fierce competition from AMD and Nvidia, and aims to improve cost-efficiency and performance.

Snap is preparing to roll out a game-changing AI model designed specifically for mobile devices, a move that will significantly elevate the Snapchat experience in the coming months. Unveiled on Tuesday, this new AI model promises to generate high-resolution images in as little as 1.4 seconds on the iPhone 16 Pro Max—an impressive feat for mobile technology.

Presentations.ai, an innovative Indian startup using AI to streamline the presentation creation process, has raised $3 million in a seed funding round led by Accel. The company’s AI-powered platform, which recently emerged from its beta phase, has already attracted millions of users worldwide and is set to scale its offerings further.

OpenAI has introduced o3-mini, its latest AI reasoning model designed to tackle technical domains with precision and speed, marking a crucial moment for the company. Launched on January 31, 2025, o3-mini expands OpenAI's o family of reasoning models, offering a blend of power and affordability aimed at improving accessibility to advanced AI.

When we talk about "tabular data," we’re referring to any type of structured information that fits neatly into rows and columns. This could be data from SQL databases, spreadsheets, CSV files, or similar formats. It’s the kind of data businesses have been using for years to track everything from sales numbers to customer behavior.

Google took a low-key approach with the debut of its next-gen AI model, Gemini 2.0 Pro Experimental, quietly slipping it into a changelog update for the Gemini chatbot app rather than making a big public announcement. This model follows the Gemini 1.5 Pro from last February and comes at a time when AI companies, especially Chinese startup DeepSeek, are making waves with models that rival or surpass those from American tech giants. DeepSeek’s success has triggered major conversations in Silicon Valley and even Washington, as the global AI competition heats up.

On Wednesday, Google introduced its highly anticipated new AI flagship model, Gemini 2.0 Pro Experimental, alongside a slew of other exciting updates in the AI space. The release also includes the company’s reasoning-focused model, Gemini 2.0 Flash Thinking, now available through the Gemini app.

As an age when AI is changing the way things work, startups and big tech companies are using its power to aid developers with enhancing coding skills and getting ready for the real world scenarios, such as for job interview. Tech assessment platform CodeSignal, with an existing platform which includes AI driven learning for technical subjects, has now dipped its toes into the soft skill training industry offering courses that help individuals enhance their skills in different leadership, communication and other essential professional abilities.

Apple CEO Tim Cook recently discussed DeepSeek’s AI models during an earnings call, highlighting how innovation that drives efficiency is key for the tech giant’s AI strategy. When asked about the potential impact of DeepSeek’s models on Apple’s margins, Cook responded, stating that efficiency is always beneficial, and DeepSeek’s models represent such innovation.

On Thursday, Ai2, a nonprofit AI research institute based in Seattle, unveiled its latest creation, the Tulu 3 405B model, claiming that it outperforms DeepSeek V3, one of the leading systems from Chinese tech company DeepSeek.

Following Amazon's recent integration of generative AI for advertisers, Google is unveiling a suite of generative AI product imagery tools for U.S. advertisers today. With the innovative AI-powered Product Studio, merchants and advertisers have the opportunity to create new product images at no cost by simply entering a description of the desired image.

Google has introduced Whisk, a novel AI tool from Google Labs that serves as an image generator. What sets Whisk apart is its unique approach to using existing images as prompts. Instead of merely replicating the original with added details, Whisk captures the "spirit" or "essence" of the image, making it a powerful tool for sparking creativity and rapid visual brainstorming. This makes it less about perfect replication and more about inspiring new ideas, which is a refreshing take on AI image generation.

Google is in the midst of a fundamental transformation, integrating artificial intelligence directly into Search. CEO Sundar Pichai, speaking during Alphabet’s latest earnings call, made it clear that this shift is just beginning. The company’s AI journey kicked off with AI Overviews, a feature that replaces traditional search results with AI-generated responses. Despite its rocky debut—marked by bizarre AI errors—Google is doubling down on AI-driven search innovation in 2025.