AI This Week

Baidu is set to release its Ernie generative AI large language model to the open-source market, marking China's most significant AI development since DeepSeek. The rollout signals a shift from Baidu’s traditional proprietary strategy and intensifies competition among leading global tech firms. Experts suggest the move could disrupt pricing dynamics, pushing competitors like OpenAI and Anthropic to reassess their gated premium models. Industry leaders are divided on whether Ernie will rival DeepSeek’s impact, but many agree its open-source nature raises the stakes for innovation and accessibility. Questions surrounding security, transparency, and geopolitical implications also loom large as Baidu positions its AI tools for widespread adoption.

A recent study from the Max Planck Institute reveals that people, especially academics and professionals, have begun adopting AI-like language patterns in their daily speech. By analyzing 280,000 academic YouTube videos, researchers identified a significant increase in the use of terms commonly generated by AI, such as “meticulous” and “adept.” This linguistic shift might erode emotional nuance and individuality, reducing speech to monotony. The study warns that overusing AI-influenced language could flatten cultural diversity and subtly alter social behaviors, like politeness in conversations. The irony? Humans created AI to sound human, but now people are mirroring AI.

Washington University School of Medicine in St. Louis has developed an AI-based brain mapping technology that has received FDA authorization for use in neurosurgery. The Cirrus Resting State fMRI Software rapidly identifies critical brain areas controlling functions like speech and movement, enhancing precision in tumor removal and epilepsy treatment. This technology, licensed to startup Sora Neuroscience, analyzes brain activity patterns in just 12 minutes of fMRI scanning. It offers broader accessibility for brain mapping, including for patients who struggle with traditional task-based fMRI

Walmart has unveiled advanced artificial intelligence tools designed to empower its workforce and improve customer interactions. Among these innovations is a real-time translation feature supporting 44 languages, which ensures clear communication between employees and shoppers. This tool is tailored to Walmart’s environment, recognizing specific terms like “Great Value,” the company’s private label. Employees will also benefit from upgraded conversational AI, capable of handling detailed questions and offering guided responses. Additionally, Walmart is rolling out augmented reality technology to assist associates in efficiently locating merchandise, focusing on apparel stocking needs. Powered by its proprietary machine learning system, Element, Walmart’s latest AI-driven initiatives aim to enhance operational efficiency across its 10,750 stores worldwide.

A new study claims artificial intelligence surpasses human emotional understanding, especially in intense situations. Researchers from the University of Geneva and the University of Bern applied emotional intelligence tests to various AI language models like ChatGPT-4 and Gemini 1.5 Flash. The AI achieved a 81% accuracy rate in selecting correct emotional responses, compared to 56% for humans. Experts caution that the tests may oversimplify human emotions, highlighting AI's pattern recognition rather than genuine emotional insight. These findings suggest AI’s potential in real-world applications, as demonstrated by Aílton, a conversational AI used by Brazilian truck drivers.

DeepL, an AI translation company, has announced a dramatic reduction in processing time for translating the entire internet: from 194 days to just 18.5 days. This achievement is made possible by Nvidia's DGX SuperPOD with DGX GB200, located in Sweden. The advanced system also allows DeepL to translate the Oxford English Dictionary in two seconds and process lengthy novels in less than 0.09 seconds. This technological leap aims to enhance DeepL's language AI platform, improve translation accuracy, and develop new features.

DARPA has launched its “Exponentiating Mathematics” (expMath) initiative, a bold move to accelerate mathematical discovery using AI. The goal: build AI “co-authors” — systems capable of breaking down complex proofs into manageable lemmas and composing novel theorems. DARPA is seeking proposals combining large language models, reinforcement learning, and program synthesis to advance symbolic reasoning in mathematics. The roadmap is modular: one track targets AI development itself, while another evaluates AI against professional-level math challenges. If successful, expMath could radically shorten research timelines, impacting cryptography, materials science, fluid dynamics, and more. The program also seeks to democratize access to advanced math tools. DARPA’s ambition is clear: redefine how humans and machines collaborate on deep mathematical inquiry.

A recent NBC News poll finds American views on artificial intelligence echo the early buzz about the internet from the 1990s. AI chat tools such as ChatGPT, Google’s Gemini and Microsoft’s Copilot become widely adopted. Nearly three-fourths of adults report having used these tools while 44% say they use them regularly. The poll draws parallels with past attitudes toward the internet, noting that many concerns remain as new challenges and benefits arise. Opinions sharply divide on AI’s promise; 44% believe AI will improve life, while 42% worry it will harm society. Many even support classroom bans as a way to better prepare students for the future.

Reddit has unveiled two AI-powered ad tools — Reddit Insights and Conversation Summary Add‑ons — built on its “Community Intelligence” engine. These tools analyze billions of comments and posts to extract real‑time trends and highlight positive user feedback directly beneath sponsored content. They aim to help advertisers craft smarter campaigns and deepen engagement by weaving authentic community voices into ads. Launched at Cannes Lions Festival 2025, they reflect a broader industry shift toward AI-driven marketing amid sluggish ad markets. The move also raises privacy questions, as users’ public input fuels ad targeting without explicit notice. These features mark a major push by Reddit to monetize its vast user-generated content while stepping into next‑gen advertising tactics.

TikTok is introducing new artificial intelligence-powered advertising features. The platform now allows marketers to generate video ads using text prompts or still images. Advertisers can upload product images or provide text descriptions, and TikTok's AI tools will create various five-second video clips suitable for ads. This move expands TikTok's AI-generated content capabilities, offering advertisers more options to create engaging video ads quickly and efficiently. The new tools aim to streamline the ad creation process and enhance marketers' ability to produce content tailored for the platform.