Google’s Gecko benchmark identifies best AI image generator
Google’s DeepMind released Gecko, a new benchmark for comprehensively evaluating AI text-to-image (T2I) models. Over the last two years, we’ve…
The mysterious ‘gpt2-chatbot’: a surprise breakthrough in AI?
A mysterious chatbot named ‘gpt2-chatbot’ has sparked immense excitement and speculation across the AI community. This chatbot, which you can…
Notable absences hit the AI Safety Summit taking place in May
Last year, the first Global AI Safety Summit was held at the UK’s historic Bletchley Park, captivating global attention. However,…
The Financial Times and OpenAI strike deal over training data
The FT is granting OpenAI access to its news archives as generative AI companies continue to secure private data sources. …
DHS launches AI safety board but big open-source names absent
United States Secretary of Homeland Security Alejandro Mayorkas announced that the DHS had established an Artificial Intelligence Safety and Security…
Gym teacher arrested for AI clone of principal’s voice
A Baltimore high school athletic director was arrested on Thursday after police said he had allegedly used AI to create…
DAI#36 – Mini models, upscale video, and AI dogfights
Welcome to our roundup of this week’s freshest AI news. This week AI took blurry videos and made them look…
Adobe’s VideoGigaGAN upscales blurry video to look 8x sharper
Adobe researchers unveiled VideoGigaGAN, a generative AI model that can upscale blurry videos into crisp, smooth video that looks up…
Microsoft launches Phi-3 Mini, a tiny but powerful LM
Microsoft launched Phi-3 Mini, a tiny language model that is part of the company’s strategy to develop lightweight, function-specific AI…
LLM agents can autonomously exploit one-day vulnerabilities
University of Illinois Urbana-Champaign (UIUC) researchers found that AI agents powered by GPT-4 can autonomously exploit cybersecurity vulnerabilities. As AI…