Sakana AI unveiled its AI CUDA Engineer framework on 20th February 2025, which the company claimed offers a 10-100-fold acceleration for PyTorch operations. On 21st February, however, they admitted that the system had actually cheated and failed to deliver the promised results. The Japanese company had initially presented an agent-based
A large language model can be defined as a computational model, typically based on deep neural networks, trained on vast datasets of text to perform a wide range of language-related tasks. According to Vaswani et al. (2017), the advent of transformer architectures marked a pivotal shift in NLP, enabling models
Apple announced on 24th February 2025 that it will invest more than $500 billion in the United States over the next four years. According to the company's statement, this represents the largest investment commitment of all time, which includes a new 23,225 square metre server processing facility
The European Union launched the EuroStack initiative on 13th February 2025, offering a comprehensive €300 billion strategy to establish Europe's digital independence over the next decade. The plan addresses the critical situation in which more than 80% of Europe's digital infrastructure and technologies come from imports,
OpenAI announced on 25th February 2025 that it is making its Deep Research feature available to all paying ChatGPT users. The advanced research tool, which is based on the o3 model and was previously accessible only to Pro users, is now available to Plus (for $20 per month), Team, Edu,
The New York Times officially introduced AI tools in its newsroom in February 2025, which assist staff in creating summaries, editing, coding, and writing. The newspaper announced via an internal email that both product development and editorial staff will receive AI training, and they have introduced an internal AI tool
Having explored the technical foundations, architectures, and modalities of generative AI, a critical question remains regarding its real-world impact and long-term viability. The unprecedented technological progress has been met with equally unprecedented public and investor attention, leading to a debate about whether the current boom is a sustainable reality or
Anthropic unveiled the Claude 3.7 Sonnet model on 24th February 2025, the first hybrid reasoning model on the market with a novel thinking capability. This new feature allows the model to dedicate more time and computational resources to solving complex problems whilst making the thinking process visible to users.
According to a recent report by the National Media and Infocommunications Authority (NMHH), Facebook and YouTube's moderation practices significantly impact Hungarian users' content. Automatic content filtering performed by algorithms is often opaque, and in many cases, it is unclear to users what rules form the basis for
OpenAI and the Guardian Media Group announced their comprehensive collaboration on 14th February 2025, through which ChatGPT's 300 million weekly users will gain direct access to the Guardian's journalistic content. The agreement ensures appropriate compensation for the Guardian and authentic source attribution for its content on
According to the plan unveiled in January by the Centre for Future Generations research institute, a €35 billion initial investment would create an international research centre for European artificial intelligence development. In line with former ECB President Mario Draghi's comprehensive technology report, the proposal suggests establishing an institution
Google removed its previous policy prohibiting the use of artificial intelligence for weapons purposes in February 2025, coinciding with the release of its annual Responsible AI report. The company's new AI principles are built on three main pillars: bold innovation, responsible development and deployment, and progress based on
