Agents Need to Learn by Watching

<!-- SC_OFF --><div class="md"><p>I'll be upfront about my endgame: I work at a large tech company, I have a niche picked out, and I'm making the move to build and sell AI agents

<!-- SC_OFF --><div class="md"><p>Gemini 3.5 Flash launched Tuesday, and Google barely framed it as a conversational product. Instead, the company focused on coding pipelines,

<!-- SC_OFF --><div class="md"><p>I think we're entering a strange phase where software is no longer helping users do tasks better, it's trying to remove the user from the task

<!-- SC_OFF --><div class="md"><p>I've been experimenting with different ways to give AI agents access to a real computer (not just code execution) and wanted to share what I've

<!-- SC_OFF --><div class="md"><p>The idea is simple, if one agent is good then multiple agents working together must be better. In practice coordination between agents is an

<!-- SC_OFF --><div class="md"><p>I’ve been thinking about a problem in current agent systems:</p> <p>Most agents are becoming very good at execution, but the decision layer

<!-- SC_OFF --><div class="md"><p>Cursor dropped their new product yesterday and its not a code editor anymore. its an agent orchestration console that happens to have an editor

<!-- SC_OFF --><div class="md"><p>Just went through a Stanford paper that tracked 51 actual AI deployments - not surveys or sentiment polls, real production systems across 41