InsightfulTechnical

19 Ways to Use Claude So Well It Feels Illegal

Sabrina Ramonov ๐Ÿ„

This transcript is a rapid-fire overview of 19 advanced Claude AI features and usage tips, ranging from token management and model switching to browser control and automated task scheduling. The content is presented as a teaser for a more detailed tutorial available via Instagram DM.

Summary

The transcript covers a fast-paced list of 19 ways to use Claude more effectively, seemingly from a short-form video. The tips span several categories of functionality.

On the thinking and model side, the speaker mentions the ability to chat while Claude is already thinking, swap models to save tokens, set how hard Claude will think, get a better model's second opinion, and access a 'smartest deep thinking mode.'

For context and token management, the tips include reducing output tokens, viewing a breakdown of current context usage, and deleting old context to stay accurate and save tokens โ€” suggesting users can actively manage session memory and cost.

In terms of productivity and automation, the speaker highlights features like suggesting ways to work 10 times faster, creating high-quality skills with benchmarks, and teaching new teammates how to use a project. There are also mentions of repeating tasks during a session while the laptop is open, running tasks automatically even when the laptop is off or while the user is asleep, and controlling Claude Code from a phone.

Finally, the transcript touches on interface and navigation features such as giving a current chat a new name, picking up where you left off, letting Claude control your web browser, going back to an earlier message, and customizing your status line with live costs and progress bars.

The video ends with a call to action, asking viewers to follow and comment 'commands' on Instagram to receive a detailed step-by-step tutorial via DM.

Key Insights

  • The speaker claims Claude can run tasks automatically even when the user's laptop is off and while they are asleep, implying a background or server-side execution capability.
  • The speaker states users can set how hard Claude will think, suggesting adjustable reasoning depth or compute levels are a user-facing control.
  • The speaker asserts that deleting old context improves accuracy and saves tokens, framing context management as a deliberate and beneficial user action.
  • The speaker highlights the ability to let Claude control your web browser, positioning browser automation as a built-in or accessible Claude feature.
  • The speaker describes the ability to get a better model's second opinion, implying users can route queries across multiple Claude models for validation or quality checks.

Topics

Claude AI features and tipsToken and context managementAutomation and task schedulingModel selection and thinking modesProductivity workflows

Full transcript available for MurmurCast members

Sign Up to Access

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.