Le projet THINK

Rethinking AI Alignment and Safety in the Age of Deep Scheming

The first post in this series introduced vector search, its relevance in today’s world, and the important metrics used to characterize it. We can achieve dramatic gains in vector search systems by improving their internal vector representations, as the majority of the search runtime is spent bringing vectors from memory to compute their similarity with the query. The focus of this post, Locally-adaptive Vector Quantization (LVQ), accelerates the search, lowers the memory footprint, and preserves the efficiency of the similarity computation.

Discover how dstack’s integration with Intel® Gaudi® AI accelerators enhances AI workflow efficiency, reduces costs, and accelerates scalable model deployment across cloud and on-prem environments.

A new product from Intel Tiber Trust Services brings zero-trust security to AI model training.

Discover how AI innovators tackled creative challenges, built multimodal applications, and won at the Advent of Multimodal AI Hackathon. Here are the highlights.

AI meets creativity! See the top innovations from the Advent of Multimodal AI Hackathon, from AI-generated manga to smart marketing tools. Read more to discover the top submissions.

Intel® Liftoff startups led the charge at the Advent of Multimodal AI Hackathon, showcasing clever AI solutions. Read more about their innovations.

ElectricSQL’s local-first approach delivers instant mobile data access without cloud storage, improving user experience and cutting costs. Read the full article to learn more.

Register for Intel AI DevSummit 2025.

From security and efficiency to testing, Intel® Liftoff Startups have jumped at the chance to explore DeepSeek’s potential. Here’s a look at the buzz, the breakthroughs, and why so many teams are betting on DeepSeek.