Lazarus AI

"The mini-labs provide a flexible test environment that delivers real measurement data," Gamella said in a statement. "They're a bit like a playground for algorithms, where researchers can test their AI beyond simulated data in a controlled and safe environment."

Another week of AI breakthroughs, backtracks, and bold predictions. If you blinked last week, here's what you missed: Musk's Grok went rogue (briefly), Amazon is trying to make Alexa relevant again, OpenAI's latest model does relationship advice,

The use of LLMs in healthcare involves sensitive patient data, making privacy a critical concern. Data leakage, unauthorized information retention, and vulnerabilities to adversarial attacks pose significant risks to patient confidentiality.

The model with extended thinking beats OpenAI's o1 and o3-mini and DeepSeek's R1 in virtually every category besides math problem-solving. In particular, Claude 3.7 Sonnet is particularly strong in coding, easily beating other models. It is also strong in graduate-level reasoning and interactions with external systems.

Figure launches Helix, an AI that allows robots to recognize language and reasoning like humans and grasp household objects without training.

Learn whether a smaller Diffbot's AI model with an innovative GraphRAG AI training technology can solve AI hallucinations for good.

Advanced AI systems are set to emerge as the solution companies have been waiting for.

The result was Athenic AI, a company that uses AI to run data analytics for enterprises across all of their data sources. Zhao, the founder and CEO, said that Athenic's products are designed to be a central nervous system of an organization's databases that can be used by anyone in the company regardless of their coding or data experience.