Modelos, agentes, investigación y negocio de la IA.
Amazon openly criticizes the AI governance model with human oversight ('human-in-the-loop'), arguing that humans are inconsistent and prone to 'normalization of deviance.' The company proposes a machine-based approach with strategic, not operational, human oversight.
Miami-based startup Subquadratic has released independent results supporting its claim of breaking quadratic attention, the mathematical bottleneck limiting language models. Its SubQ model achieves a 12 million token context window with sublinear scaling, a breakthrough that could redefine long-text processing in AI.
Norway becomes the first country to ban generative AI in primary schools, citing risks to critical thinking and minors' privacy. The decision, affecting hundreds of thousands of students, could set a precedent for other nations.
Nvidia has injected $6 billion into three optical component manufacturers (Coherent, Lumentum, and Marvell) to prepare the supply chain for massive demand of optical interconnects in large-scale AI systems. The move responds to the need to overcome copper limitations when GPU clusters reach thousands of units.
Apple faces growing investor skepticism after a WWDC that failed to convince that its AI strategy is close to driving a new upgrade cycle. Shares fell in their worst week since February, reflecting palpable fatigue with the company's promises.
Allbirds, the sustainable footwear company once valued at $4.1 billion in 2021, has rebranded as Smartbird and announced a radical pivot toward artificial intelligence infrastructure. Shares rose more than 50% following the announcement. We analyze what this move means and its implications for the market.
The US government ordered Anthropic to cut off access to its most advanced models for foreign citizens, triggering a global suspension. Mistral AI seized the opportunity to denounce American control and advocate for sovereign AI in Europe.
A study from the Wharton School reveals that many AI users are adopting 'cognitive surrender': passively accepting chatbot responses without critical evaluation. This phenomenon, named by researchers Steven Shaw and Gideon Nave, raises serious concerns about loss of autonomy and the erosion of critical thinking in the age of artificial intelligence.
OpenAI has announced an improvement to ChatGPT's memory, allowing it to remember preferences and context. However, tests reveal that the feature can generate erroneous profiles, rely on outdated assumptions, and worsen response quality, posing risks of bias and misinformation.
A new artificial intelligence approach applies language models to smells to diagnose diseases through breath. Startup Ainos, in collaboration with National Taiwan University, will test its 'AI Nose' platform on patients with dyspnea, opening the door to non-invasive and rapid diagnoses in emergency rooms.
NeuralTrust, a Barcelona-based startup, has raised €17.2 million in a seed round for its security and governance platform for enterprise AI agents. The firm claims it is the largest seed funding round in cybersecurity in the European Union to date.
Faced with rising AI API fees and limitations on building new data centers, a trend is emerging of migrating AI workloads from the cloud to local environments. Users report significant savings by processing millions of tokens daily with modest hardware, raising questions about the future of the cloud inference market.