Think Different by Jakub Jirák: posts, page 55 of 143
-
The Truth About Batteries: Why Phone Battery Life Hasn't Really Improved in Years
Explore why smartphone battery life has stagnated despite decades of innovation. Understand the physics of lithium-ion limits, the trade-offs.
-
The Hidden Math That Explains Why AI Keeps Getting Smarter
Neural scaling laws reveal a predictable mathematical relationship between compute, data, and AI capability that reshapes how we think about AI progress.
-
Apple MLX Is the Most Underrated AI Framework Right Now
A practical guide to Apple MLX and mlx-lm on Apple Silicon. Why MLX beats llama.cpp for some workloads, the MLX community models on Hugging Face, and LoRA fine-tuning a local LLM on a MacBook in one afternoon.
-
Agentic coding: Difference between an agent and a slow script
A field-tested take on the difference between an agent and a slow script with Agentic coding: what it rewards, where it breaks, and how to keep the workflow honest.
-
What a Typical Work Day Will Look Like in 2030
Explore a detailed scenario of the 2030 work day. From AI assistants handling routine tasks to new meeting formats and productivity rhythms, discover how.
-
Why Open Source AI Will Win Even If It Loses
Meta gave away what OpenAI charges for, and the open ecosystem is building something no single company can match. Here's why that matters more than benchmark scores.
-
The MacBook Pro Setting That Doubles Your AI Performance
How to raise the macOS GPU memory limit with sudo sysctl iogpu.wired_limit_mb to run bigger local LLMs on a MacBook Pro. Safe values per RAM tier, persistence setup, High Power Mode, and thermal tips.
-
MCP servers: Exposing read-only resources before you expose write actions
A field-tested take on exposing read-only resources before you expose write actions with MCP servers: what it rewards, where it breaks, and how to keep the workflow honest.
-
The Chip That Started the Second Cold War
The US export ban on advanced semiconductors to China triggered a geopolitical restructuring that rivals the original Cold War in its global consequences.
-
I Replaced My $20 AI Subscriptions With One Mac Studio
Cost-benefit analysis of replacing ChatGPT-style subscriptions with local LLMs on a Mac Studio. Ollama plus Open WebUI setup, what local models can and cannot replace, and the real break-even math.
-
AI coding workflows: Small reversible steps instead of one heroic prompt
A field-tested take on small reversible steps instead of one heroic prompt with AI coding workflows: what it rewards, where it breaks, and how to keep the workflow honest.
