After two years of breathless predictions about AI transformation, there remains a stark divide between promise and practice. Major tech companies continue their massive infrastructure investments – with capital expenditures approaching 30% of revenues – while many enterprise clients struggle to demonstrate meaningful returns. Recent data shows 42% of companies abandoning most of their generativeContinue reading “Workflow, Not Wizardry: The Real Levers of AI Success at Work”
Category Archives: Uncategorized
🚨 New Data Reveals Why Most Gen-AI Pilots Fail
Subscribe • Previous Issues Workflow, Not Wizardry: The Real Levers of AI Success at Work After two years of breathless predictions about AI transformation, there remains a stark divide between promise and practice. Major tech companies continue their massive infrastructure investments – with capital expenditures approaching 30% of revenues – while many enterprise clients struggle to demonstrateContinue reading “🚨 New Data Reveals Why Most Gen-AI Pilots Fail”
Open-source AI that pays for itself: Block’s Vision for AI Integration
Just when I think I’ve grasped the full landscape of AI coding assistants, another compelling tool I’d never encountered invariably surfaces. codename goose (hereafter “Goose”), an open-source agent used weekly by 5,000 Block employees, shows what happens when you give an LLM a toolbox. Built at Block (formerly Square) and released under an MIT licenceContinue reading “Open-source AI that pays for itself: Block’s Vision for AI Integration”
Can a single agent automate 90% of your code fixes? Block thinks so
Subscribe • Previous Issues Open-source AI that pays for itself: Block’s Vision for AI Integration Just when I think I’ve grasped the full landscape of AI coding assistants, another compelling tool I’d never encountered invariably surface. codename goose (hereafter “Goose”), an open-source agent used weekly by 5,000 Block employees, shows what happens when you give an LLMContinue reading “Can a single agent automate 90% of your code fixes? Block thinks so”
The Protocol Foundation: Building Enterprise-Ready AI Agent Systems
I often hear a recurring concern from teams working with AI: will our agents truly integrate with our existing systems, or are we facing an endless cycle of custom wiring for each new tool? Agentic AI protocols are emerging as a potential solution. These protocols provide the essential “plumbing” that could transform a foundation modelContinue reading “The Protocol Foundation: Building Enterprise-Ready AI Agent Systems”
Tired of Custom AI Wiring? There’s a Better Way
Subscribe • Previous Issues The Protocol Foundation: Building Enterprise-Ready AI Agent Systems I often hear a recurring concern from teams working with AI: will our agents truly integrate with our existing systems, or are we facing an endless cycle of custom wiring for each new tool? Agentic AI protocols are emerging as a potential solution. These protocolsContinue reading “Tired of Custom AI Wiring? There’s a Better Way”
Beyond Open Weights: The Path to Unconditionally Open AI
While I routinely work with both proprietary LLMs and open-weights models, my heart lies with models that are open in the fullest sense. Very early on, I noted that for foundation models, ‘open’ must comprehensively cover not just weights but also data, code, and the detailed recipes crucial for genuine reproducibility. Measured against this standard,Continue reading “Beyond Open Weights: The Path to Unconditionally Open AI”
Why this AI veteran left Google to make models ‘unconditionally open’
Subscribe • Previous Issues Beyond Open Weights: The Path to Unconditionally Open AI While I routinely work with both proprietary LLMs and open-weights models, my heart lies with models that are open in the fullest sense. Very early on, I noted that for foundation models, ‘open’ must comprehensively cover not just weights but also data, code, andContinue reading “Why this AI veteran left Google to make models ‘unconditionally open’”
Claude Opus 4 and Claude Sonnet 4: Cheat Sheet
Claude Opus 4 and Claude Sonnet 4 are Anthropic’s latest hybrid-reasoning language models. The system card explains how the models were trained on a blend of public web data, opted-in user content and proprietary sources, and how they can switch between a quick default mode and a slower “extended thinking” mode for harder problems. TheContinue reading “Claude Opus 4 and Claude Sonnet 4: Cheat Sheet”
Apple’s AI: Efficiency, Privacy, and Seamless Integration
Apple’s success has been built upon a meticulous fusion of hardware, software, and services, consistently shaping how people interact with technology while championing user privacy. However, the recent explosion in artificial intelligence, particularly generative AI, presents a new paradigm. While the company is often perceived as playing catch-up to rivals who have rapidly deployed high-profileContinue reading “Apple’s AI: Efficiency, Privacy, and Seamless Integration”