MIT Team Speeds Up AI-Based Particle Size Estimator for Drug Manufacturing
MIT researchers use AI to estimate particle size 60x faster in pharma manufacturing, reducing analysis time from 15 to 0.25 seconds.
MIT researchers use AI to estimate particle size 60x faster in pharma manufacturing, reducing analysis time from 15 to 0.25 seconds.
Mercado Libre launches Verdi, an AI platform using GPT-4o, handling 10% of customer service and potentially supporting 9,000 human operators.
Manhattan Declaration proposes 10 principles for global AI governance. Experts emphasise cooperation, risk assessment, and responsible development.
Llama 3.2 introduces vision models and lightweight text models for edge devices. New 128K token context length enables advanced on-device AI tasks.
ChatGPT users gain access to Italian content from GEDI publications. Partnership aims to improve relevance and accuracy for Italian users beyond ChatGPT.
Microsoft enhances AI security with new features: proactive risk assessments, real-time hallucination correction, and confidential inferencing.
Stanford's "Digitalist Papers" explore AI's impact on democracy, featuring 19 experts' essays on governance strategies in the AI era.
Researchers propose labelling AI systems in healthcare with usage info, including training time, data origin, intended use, biases, and warnings.
Tech giants' retreat from OpenAI's board reflects growing regulatory concerns over AI partnerships and their impact on market competition.
SAG-AFTRA has taken a step forward in protecting the rights of artists. On the 10th of April, the union, representing 160,000 media professionals, has reached a tentative agreement with major record labels that includes ground breaking protections against the use of AI in the music industry.
Rakuten partners with OpenAI to enhance services using AI. Leverages data from 1.8B users across 70 platforms. Improves customer service, shopping experience, and B2B consulting. Plans for voice and vision AI capabilities.
NVIDIA's AI tech transforms STEM education globally. GPU-accelerated curriculum in 5,000+ schools prepares students for AI-driven jobs. Offers RTX AI Toolkit, RAPIDS libraries, robotics support, and remote learning tools.
Mistral AI enables model customisation via La Plateforme and introduces AI Agents, allowing developers to create complex, shareable workflows.
The guide suggests fine-tuning, particularly parameter-efficient fine-tuning, as a more viable approach for smaller teams with limited resources compared to pre-training methods.
The guide identifies five scenarios where fine-tuning excels: customising tone and format, improving accuracy, addressing niche domains, reducing costs via distillation, and developing new abilities.
Meta AI's guide emphasises dataset quality for fine-tuning LLMs, suggesting small high-quality datasets often outperform larger low-quality ones. It compares full fine-tuning and PEFT techniques.