Author: admin

AI dev platform Hugging Face continued its push into robotics on Thursday with the release of two new humanoid robots. The company announced a pair of open source robots, HopeJR and Reachy Mini. HopeJR is a full-size humanoid robot that has 66 actuated degrees of freedom, or 66 independent movements, including the ability to walk and move its arms. Reachy Mini is a desktop unit that can move its head, talk, listen, and be used to test AI apps. Hugging Face doesn’t have an exact timeline for shipping these robots. The company’s co-founder and CEO, Clem Delangue, told TechCrunch over…

Black Forest Labs, the AI startup whose models once powered the image generation features of X’s Grok chatbot, on Thursday released a new suite of image-generating models — some of which can both create and edit pics. The most capable of the models in the new family, called Flux.1 Kontext, can be prompted with text and, optionally, a reference image to create new images, writes Black Forest Labs in a blog post. “The Flux.1 Kontext models deliver state-of-the-art image generation results with strong prompt following, photorealistic rendering, and competitive typography — all at inference speeds up to 8x faster than…

Delaware’s attorney general is hiring an investment bank to advise on OpenAI’s for-profit conversion, the Wall Street Journal reported on Wednesday. The independent evaluation could prolong the transition, or gum up OpenAI’s plans even further. OpenAI’s conversion will help it attract new investment and eventually go public, but first, the startup needs to secure approvals from state regulators. Although OpenAI and Microsoft have hired their own investment banks to advise on the deal, Delaware’s attorney general is now seeking an independent evaluation of the equity that OpenAI’s nonprofit will receive. Regulators seem to be zeroing in on a key factor…

DeepSeek, a China-based artificial intelligence company, has announced an upgrade to its AI chatbot, saying it can now offer enhanced overall logic, mathematics and programming with a reduced hallucination rate.According to DeepSeek, the upgraded model — DeepSeek-R1-0528 — has “significantly improved its depth of reasoning and inference capabilities.” The startup said the model’s overall performance is now “approaching that of leading models, such as O3 and Gemini 2.5 Pro.”Performance comparison of language models across six benchmarks. Source: DeepSeekDeepSeek’s debut of its R1 chatbot in January sent shockwaves through the AI industry and further established China as an AI force. The…

Check out the companies making headlines in Thursday trading. Salesforce — The customer relations management software maker’s shares traded 5% lower on Thursday after the company posted its latest quarterly results. RBC Capital Markets downgraded the stock, citing execution risks if the company continues acquiring. Salesforce did beat fiscal first-quarter estimates and raised its full-year outlook, and posted results a day after announcing plans to acquire data management company Informatica. C3.ai — Shares of the enterprise artificial intelligence company surged 23% after C3.ai reported strong results for its fiscal fourth quarter. The company posted a narrower-than-expected loss of 16 cents per…

Perplexity, the AI-powered search engine gunning for Google, on Thursday released Perplexity Labs, a tool for subscribers to Perplexity’s $20-per-month Pro plan that can craft reports, spreadsheets, dashboards, and more. Perplexity Labs is available on the web, iOS, and Android, and coming soon to Perplexity’s apps for Mac and Windows. “Perplexity Labs can help you complete a variety of work and personal projects,” Perplexity explains in a blog post. “Labs is designed to invest more time — 10 minutes or longer — and leverage additional tools [to accomplish tasks], such as advanced file generation and mini-app creation.” Labs, which arrives the…

Household brand Nike is not setting itself up for a successful turnaround, according to Josh Brown, CEO of Ritholtz Wealth Management and CNBC PRO contributor . The widely followed investor believes that the iconic footwear brand isn’t keeping up with the fierce competition as its celebrity spokespeople age out of popularity. The stock has fallen another 19% this year following a 30% loss in 2024. “The superstars that are all aging. LeBron James is in his 40s. Michael Jordan is 30 years retired,” Brown said on CNBC’s ” Halftime Report .” “I don’t even know what we do with something…

DeepSeek’s updated R1 reasoning AI model might be getting the bulk of the AI community’s attention this week. But the Chinese AI lab also released a smaller, “distilled” version of its new R1, DeepSeek-R1-0528-Qwen3-8B, that DeepSeek claims beats comparably sized models on certain benchmarks. The smaller updated R1, which was built using the Qwen3-8B model Alibaba launched in May as a foundation, performs better than Google’s Gemini 2.5 Flash on AIME 2025, a collection of challenging math questions. DeepSeek-R1-0528-Qwen3-8B also nearly matches Microsoft’s recently released Phi 4 reasoning plus model on another math skills test, HMMT. So-called distilled models like…

CNN  —  Former Israeli prime minister Ehud Olmert has blasted the country’s political leadership and the conduct of its military, telling CNN he is no longer able to defend Israel against accusations of war crimes. Olmert, who led the country from 2006-2009, pointed to Israel’s 11-week blockade of humanitarian aid into Gaza and the soaring number of Palestinians killed. “What is it if not a war crime?” he asked rhetorically in an interview with CNN. He said Prime Minister Benjamin Netanyahu and far-right members of his government are “committing actions which can’t be interpreted any other way.” Since the start…

Chinese AI startup DeepSeek’s newest AI model, an updated version of the company’s R1 reasoning model, achieves impressive scores on benchmarks for coding, math, and general knowledge, nearly surpassing OpenAI’s flagship o3. But the upgraded R1, also known as “R1-0528,” might also be less willing to answer contentious questions, in particular questions about topics the Chinese government considers to be controversial. That’s according to testing conducted by the pseudonymous developer behind SpeechMap, a platform to compare how different models treat sensitive and controversial subjects. The developer, who goes by the username “xlr8harder” on X, claims that R1-0528 is “substantially” less permissive…