AI and agents
Using, building, and evaluating AI tools, agents, and models.
I've run an OpenClaw agent every day since January and spent too much time babysitting it. Five days ago I moved that workload to Grok Bot and logged every job. 67 jobs, 17 interventions, three things I had to fix before I could trust it, and a short list that stays on my Mac. Continue reading →
Xcode has allowed development installs for years. AI changed how much work it takes to get a tested native app onto your phone. Continue reading →
Poland never dubbed films with a cast of actors. One man read every line, flat and unbothered, over the quieted original. ElevenLabs exists partly because its founders grew up with that voice. So I rebuilt him with their own technology: a site where any video comes back with a deadpan Polish lektor. This is how it works, including the timing algorithm that turned out to be the actual craft. Continue reading →
Your OpenClaw agent works great for 20 minutes, then silently loses its instructions. The root cause is compaction, and most users don't understand it. This guide explains the full memory lifecycle and gives you a file architecture that prevents the forgetting. Continue reading →
Most OpenClaw content right now is first-week impressions or setup tutorials. Nobody can tell you what happens after the first month because they haven't been there yet. I have. Every single day for 50+ days. Here are 20 real workflows I use daily, what broke along the way, and my honest assessment of whether it's worth your time. Continue reading →
If you're running OpenClaw, there's a good chance you're burning money without realizing it. By default, everything goes to your primary model - heartbeats, sub-agents, simple queries. This guide shows you how to cut costs by 50-80% with one config change, without losing quality on the tasks that actually matter. Continue reading →
Most developers use AI to code faster today, but they're missing the bigger opportunity. Compound Engineering is a workflow where every code review teaches the system something new, every bug fix becomes a rule it won't break again, and the knowledge compounds over time. Continue reading →
Go from a raw Excel file with half a million rows to a finished report with KPIs, charts, and cohort analysis - step by step, with real prompts you can copy. Continue reading →
A personal AI assistant that runs on a $5/month server, lives in your messaging apps, remembers everything, and actually reaches out to help you. Here's how to set it up. Continue reading →
The spec workflow that went viral: instead of telling Claude what to build, you let Claude interview you first. Here's how it works, when to use it, and how to turn it into a one-command shortcut. Continue reading →
A daily learning journey of learning how to build a large language model from scratch. Understanding transformers, attention mechanisms, GPT architecture, training process. Documenting insights, concepts and implementations as I learn how LLMs actually work under the hood. Continue reading →
Finally, an AI image generator that doesn't give up after one try. Because your creative vision deserves better than digital dice rolls. Continue reading →
My month-long exploration of AI agents, where I built practical solutions from simple chatbots to email automation. Learn when to use AI versus traditional automation, and follow my recommended learning path for AI development beginners. Continue reading →
A free Chrome extension that allows you to quickly and easily summarize articles on any website, with one click or one keyboard shortcut, you'll get a concise summary in just a few seconds. Continue reading →
If you like what you see, you'll find more stuff like this on my Twitter.














