Self-hosting large language models is a hardware-intensive, patience-demanding endeavor requiring at least 16GB of VRAM for even modest models, with operational friction from RAG pipelines, prompt template mismatches, and latency issues making it far from a seamless API replacement.
OpenAI's GPT-4.5 has been officially banned from spontaneously discussing goblins, gremlins, raccoons, and trolls after the model repeatedly brought up such creatures unprompted, with the quirky prohibition going viral and even CEO Sam Altman leaning into the joke as fans build 'goblin mode' workarounds.
Google DeepMind launches ProEval, a free open-source tool that slashes generative AI evaluation costs by up to 100x using Bayesian Quadrature techniques, achieving ±1% accuracy with a fraction of typical samples while proactively identifying model failure patterns across major benchmarks.
A groundbreaking new AI reasoning framework called LaDiR is outperforming all existing models across math and planning benchmarks by combining latent diffusion models with continuous representations to generate more accurate, diverse, and interpretable reasoning trajectories.
Nvidia launches Nemotron 3 Nano Omni, a powerful open multimodal AI model unifying vision, speech, and language in one system, topping six leaderboards and delivering up to 9x faster throughput than competing models, now available on Hugging Face and OpenRouter.
Amazon launches Amazon Quick, a free AI work assistant with deep integration across Microsoft 365, Google Workspace, Salesforce, and Zoom, offering enterprise-grade automation for drafting emails, updating tickets, and creating presentations — trusted by major companies like Southwest Airlines, BMW, and the NFL.
Perplexity launches a massive enterprise expansion of its AI agent platform at Ask NYC, introducing 50+ prebuilt workflows, Databricks and Snowflake integrations, financial data connections via Morningstar and PitchBook, and a 1Password integration — all while claiming $2.8 billion in labor-equivalent work completed since launch.
AI startup Rogo Technologies hits a $2 billion valuation after two former Wall Street bankers build a platform that automates the tedious grunt work of investment banking, including spreadsheet crunching and slide deck formatting.
Biohub launches its $500 million Virtual Biology Initiative, a five-year effort to build open, AI-powered predictive models of human cells, partnering with major institutions like the Broad Institute and Allen Institute, with NVIDIA providing computing power, and all generated data made freely available to accelerate disease prevention and cures.
The Zuckerberg Chan Biohub is committing $500 million over five years to build AI-powered simulations of the human body, partnering with Nvidia and top research institutions in an ambitious push to ultimately cure all human disease.