PrismML Releases Bonsai 27B, an Open Reasoning Model for iPhone
PrismML has released Bonsai 27B, a 27-billion-parameter AI model that runs on an iPhone without sacrificing reasoning or agent capabilities, according to a CNBC report.
Archive
Browse all published articles.
ARCHIVE
-
-
Anthropic and Blackstone Launch Ode, AI Implementation Firm
Anthropic and Blackstone have launched Ode, a $1.5 billion AI implementation company, aiming to help businesses adopt AI technologies effectively.
-
Apple Intelligence Approved for Launch in China with Alibaba's Qwen AI
Apple Intelligence, Apple's generative AI offering, has received approval from China's regulator to launch in the country, following an integration deal with Alibaba's Qwen AI model.
-
AMD Analyzes FP8 KV Cache Performance on MI308X
AMD's case study shows FP8 KV cache on MI308X increased GPU time by 0.08% despite a 34% faster MLA decode kernel.
-
Cohere Explains Total Cost of AI Ownership
Cohere highlights the rising costs of AI ownership as global spending on AI is projected to reach $2.52 trillion by 2026, driven by infrastructure needs.
-
Apple's FaceID Creator Launches AI Model for Brain Diagnostics
Hemispheric, founded by Apple's FaceID co-inventor, raised $52 million to develop an AI model that uses brain data from 100,000 volunteers to diagnose cognitive disorders.
-
Vint Cerf Advises on Open-Internet AI Agent Standards
Vint Cerf, a founding father of the internet, is helping shape open standards for AI agents, aiming to create a shared framework for accountability and identification.
-
Rime Secures $24M Series A to Power Enterprise Voice AI
Voice AI startup Rime raised $24 million in Series A funding to enhance enterprise customer call solutions. The round was led by M13 Ventures with participation from Twilio Ventures and other existing investors.
-
Hugging Face Evaluates Open-Source AI Models for Swiss Legal Tasks
Hugging Face evaluated 13 open models on three Swiss legal benchmarks with 55,861 scored samples each, showing the top five models differ by only 2.1 points.
-
Meta Employees Sue Over AI-Driven Layoffs
Former and current Meta employees allege discriminatory layoffs were driven by AI systems that targeted disabled workers and those on leave. The lawsuit claims 8,000 employees were cut in May.
-
OpenAI's Codex Now Encrypts AI Agent Instructions
OpenAI's Codex now encrypts internal instructions between AI agents, obscuring task delegation from developers. The change affects larger GPT-5.6 models starting in early June.
-
OpenAI Employees Fund Guardrails Alliance Against Pro-AI PAC
OpenAI employees donated over $215,000 to Guardrails Alliance, a super PAC seeking stricter AI regulations, as tensions rise within the company over its political ties.