Author: Advanced AI Editor

Posted Sep 5, 2025 at 4:20 PM UTCOpenAI is acqui-hiring the team behind Alex, a Y-combinator backed startup.They’ll be joining OpenAI’s Codex team, which is working on the company’s AI coding assistant. “Building a ‘Cursor for Xcode’ sounded crazy, but we managed to do it anyway,” Alex founder Daniel Edrisian wrote in a post on X, adding, “And, over time, we built the best coding agent for iOS & MacOS apps … It is an honor to continue that work at a much bigger scale at OpenAI.”Follow topics and authors from this story to see more like this in your…

Authors revealed today that Anthropic agreed to pay $1.5 billion and destroy all copies of the books the AI company pirated to train its artificial intelligence models. In a press release provided to Ars, the authors confirmed that the settlement is “believed to be the largest publicly reported recovery in the history of US copyright litigation.” Covering 500,000 works that Anthropic pirated for AI training, if a court approves the settlement, each author will receive $3,000 per work that Anthropic stole. “Depending on the number of claims submitted, the final figure per work could be higher,” the press release noted.…

Want to keep track of the largest startup funding deals in 2025 with our curated list of $100 million-plus venture deals to U.S.-based companies? Check out The Crunchbase Megadeals Board. This is a weekly feature that runs down the week’s top 10 announced funding rounds in the U.S. Check out last week’s biggest funding rounds here. Investor enthusiasm for AI once again drove the bulk of funding for the week’s largest rounds, led by Anthropic’s $13 billion mega-financing. However, companies across a number of industries attracted large investments, including biotech, security and medical devices. 1. Anthropic, $13B, generative AI: San…

Dot, an AI companion app that aimed to be a friend and confidante, is shutting down, the company announced on Friday. On a message published on its website, the startup behind Dot, New Computer, said that the product will remain operational until October 5, giving users time to download their data. Launched in 2024 by co-founders Sam Whitmore and former Apple designer Jason Yuan, Dot waded into what’s now become a more controversial area for AI chatbots. The app they created was described as an AI “friend and companion,” which would become more personalized to you and your interests over…

Midjourney could slow down the cycle and “minimize” these allegedly infringing outputs, if it cannot automatically block them all, WB suggested. But instead, “Midjourney has made a calculated and profit-driven decision to offer zero protection for copyright owners even though Midjourney knows about the breathtaking scope of its piracy and copyright infringement,” WB alleged. Fearing a supposed scheme to replace WB in the market by stealing its best-known characters, WB accused Midjourney of willfully allowing WB characters to be generated in order to “generate more money for Midjourney” to potentially compete in streaming markets. Midjourney will remove protections “on a…

Fadaee takes over for Sara Hooker to lead the AI company’s research lab. Toronto-based artificial intelligence (AI) scaleup Cohere has promoted staff research scientist Marzieh Fadaee to head of its non-profit research division, Cohere Labs, replacing outgoing leader Sara Hooker. “It’s an honor and a responsibility to lead such an extraordinary group of researchers pushing the boundaries of AI research,” Fadaee wrote in LinkedIn and X posts announcing the news. Fadaee thanked Hooker for shaping Cohere Labs and said she looks forward to working with Joelle Pineau, Cohere’s new chief AI officer, to write the AI research lab’s “next chapter.”…

(Dave Hoeek/Shutterstock) Gartner says the average company spent around $1.9 million on GenAI last year, yet fewer than 30% of AI leaders think their CEOs are satisfied with the results. That gap between spending and satisfaction is concerning. After a stretch of buzz and experimentation, business leaders are moving past flashy demos and proof-of-concept hype. They’re asking harder questions now. What can AI really do inside a complex business? What works at scale, and what breaks when real-world systems get involved? You can see that shift clearly in Gartner’s latest Hype Cycles for Artificial Intelligence and Generative AI. These reports…

This article was written byFollowJulian Lin is a financial analyst. He finds undervalued companies with secular growth that appreciate over time. His approach is to look for companies with strong balance sheets and management teams in sectors with long growth runways. Julian is the leader of the investing group Best Of Breed Growth Stocks where he only shares positions in stocks which have a large probability of delivering large alpha relative to the S&P 500. He also combines growth-oriented principles with strict valuation hurdles to add an additional layer to the conventional margin of safety. Features include: exclusive access to…

We present NER Retriever, a zero-shot retrieval framework for ad-hoc Named Entity Retrieval, a variant of Named Entity Recognition (NER), where the types of interest are not provided in advance, and a user-defined type description is used to retrieve documents mentioning entities of that type. Instead of relying on fixed schemas or fine-tuned models, our method builds on internal representations of large language models (LLMs) to embed both entity mentions and user-provided open-ended type descriptions into a shared semantic space. We show that internal representations, specifically the value vectors from mid-layer transformer blocks, encode fine-grained type information more effectively than…

This post was written with Mohamed Hossam of Brightskies. Research universities engaged in large-scale AI and high-performance computing (HPC) often face significant infrastructure challenges that impede innovation and delay research outcomes. Traditional on-premises HPC clusters come with long GPU procurement cycles, rigid scaling limits, and complex maintenance requirements. These obstacles restrict researchers’ ability to iterate quickly on AI workloads such as natural language processing (NLP), computer vision, and foundation model (FM) training. Amazon SageMaker HyperPod alleviates the undifferentiated heavy lifting involved in building AI models. It helps quickly scale model development tasks such as training, fine-tuning, or inference across a…