~ai-venturebeat | Bookmarks (565)

It is a well-known fact that different model families can use different tokenizers. However, there has been limited analysis on how the process of “tokenization” itself varies across these tokenizers. Do all tokenizers result in the same number of tokens for a given input text? If not, how different are the generated tokens? How significant […]

Astronomer secures $93 million in Series D funding to solve the AI implementation gap through data orchestration, helping enterprises streamline complex workflows and operationalize AI initiatives at scale.

The release demonstrates that with carefully curated data and training techniques, small models can deliver strong reasoning performance.

Salesforce unveils groundbreaking AI research tackling "jagged intelligence," introducing new benchmarks, models, and guardrails to make enterprise AI agents more intelligent, trusted, and consistently reliable for business use.

UiPath's agent orchestration layer Maestro moves prompts through three layers: the agent, a human and the robotic process automation system.

AI visionaries predict an 'Era of Experience' where AI learns autonomously, and it will have important implications for application design.

How University of Texas Medical Branch is using AI to identify patients at high cardiovascular risk, flag for stroke and catch 'basic stuff.'

Brooklyn-based Structify emerges from stealth with $4.1 million in seed funding to transform how businesses prepare data for AI, promising to save data scientists from the task that consumes 80% of their time.

Mastercard is working with AI companies and banks to allow AI platforms and agents to facilitate transactions.

Meta partners with Cerebras to launch its new Llama API, offering developers AI inference speeds up to 18 times faster than traditional GPU solutions, challenging OpenAI and Google in the fast-growing AI services market.

This mainstream exposure will likely accelerate a shift in what people expect not just from consumer apps, but from workplaces...

Tripp launched Kōkua AI, a mental wellness guide designed to deliver real-time, personalized emotional support across multiple platforms.

xMEMS Labs, a pioneer of MEMS-based chips, announced that its innovative µCooling fan-on-a-chip tech will be expanded to AI data centers.

Qwen3’s open-weight release under an accessible license marks an important milestone, lowering barriers for developers and organizations.

Crucially, the turbulence also nudges many organizations to explore open-source models they can host, monitor, and fine-tune themselves.

The LOKA protocol, a proposed standard for AI agents from Carnegie Mellon University researchers, will give identities and intentions to agents.

Writer unveils Palmyra X5: The enterprise AI model that processes 1,500 pages at once, costs 75% less than GPT-4, and enables affordable autonomous agents for businesses seeking automation ROI.

RAG is supposed to make enterprise AI more accurate, but it could potentially also make it less safe according to new research.

Metrics are critical for determining AI product performance. But where to begin? Here's a framework to apply across various use cases.

How did DeepSeek attain such cost-savings while American companies could not? Let's dive into the technical details.

Hyena Edge’s success positions Liquid AI as one of the emerging players to watch in the evolving AI model landscape.