Topic: model development
-
Google Targets Gemini 4 Launch Before Year-End
Google DeepMind is accelerating the launch of Gemini 4, which has entered its post-training phase and is currently undergoing internal testing with its coding assistant. New lab leader Koray Kavukcuoglu aims to release an early output as soon as possible following a recent leadership transition t...
Read More » -
Fish Audio secures $50M seed for AI voice models targeting creators and enterprises
Fish Audio offers a library of over 15,000 natural language controls for voice generation, serving both creative professionals and enterprises, and has attracted 8 million users with $21 million in annual recurring revenue. The startup has secured $50 million in seed funding and recently automate...
Read More » -
OpenAI files for IPO, beating Anthropic in race to go public
OpenAI is set to confidentially file its IPO prospectus this week, with Goldman Sachs and Morgan Stanley as underwriters, marking a strategic shift from private innovation to public accountability. This move gives OpenAI a decisive lead in the AI race to go public, ahead of rival Anthropic, and p...
Read More » -
SpaceXAI staff exodus continues after merger with Musk firm
Over 50 researchers and engineers have left SpaceXAI since February, affecting key teams like coding and the Grok voice assistant, following the merger of xAI into SpaceX. Competitors like Meta and Mira Murati's Thinking Machine Labs have hired at least 18 former employees, while the core pre-tra...
Read More » -
OpenAI Expands London Office in Major Growth Move
OpenAI is significantly expanding its London office into its largest research hub outside the U.S., intensifying the competition for top AI talent with established players like Google DeepMind. The expansion focuses on recruiting researchers from UK universities and delegating core responsibiliti...
Read More » -
Grok AI Now Excels at Baldur's Gate Questions
Elon Musk's xAI delayed a model launch to specifically improve its Grok chatbot's ability to answer complex questions about the video game Baldur's Gate, reallocating engineering talent for this purpose. In a comparison test, Grok provided competent, well-informed advice for the game but often us...
Read More » -
Runway AI Raises $315M at $5.3B Valuation for Advanced World Models
Runway AI raised $315 million, doubling its valuation to $5.3 billion, with funds targeted at developing advanced "world models" for simulating and predicting real-world environments. The company is expanding beyond its creative industry roots, seeing growing adoption in gaming and robotics, and ...
Read More » -
Nearly Half of xAI's Founding Team Departs
xAI is facing significant talent attrition, with five of its twelve original co-founders having departed, raising concerns about stability ahead of a major public offering. The exits are attributed to factors including the demanding work environment, impending financial windfalls from the IPO, an...
Read More » -
OpenAI's GPT-5.2 "Screwed Up," Admits Sam Altman
OpenAI CEO Sam Altman acknowledged that the new GPT-5.2 model has significant issues with writing quality, a regression from the previous GPT-4.5 version. The decline in writing fluency resulted from a strategic focus on enhancing technical and reasoning capabilities, such as coding and complex p...
Read More » -
Europe's Race to Build Its Own DeepSeek
Europe is intensifying efforts to build a sovereign AI capability, driven by geopolitical tensions and a desire to reduce strategic dependency on dominant U.S. technology firms. A distinct European strategy is emerging, centered on open-source models and collaborative development to foster a dece...
Read More » -
2025: The Year AI Faced Its Vibe Check
2025 marked a turning point for AI, shifting from hype to scrutiny over sustainable business models, innovation pace, and societal impact, despite continued massive investment. The industry faced a reality check with more incremental model improvements, a disruptive focus on cost-effective develo...
Read More »