Central Theme
This article provides a practical guide on how to effectively use today’s leading AI systems, arguing that the focus should shift from finding the single “best model” to mastering the features within the best overall systems. The author, Ethan Mollick, aims to help users move beyond simple, Google-like queries to become power users who leverage AI for complex, high-stakes work.
Key Points & Findings
- Which AI to Use: For serious use, choose one of the three main systems: OpenAI’s ChatGPT, Google’s Gemini, or Anthropic’s Claude. A paid subscription (~$20/month) is necessary to access the full suite of essential features.
- Model Selection is Crucial: Each system offers multiple models. Users must manually switch from the fast, default models (e.g., GPT-4o, Gemini Flash) to the more powerful models (e.g., o3, Gemini Pro, Claude Opus) for any serious analysis, writing, or research to ensure higher quality results.
- Leverage “Deep Research”: This is a key, underutilized feature that generates high-quality, cited reports on complex topics. It is far more accurate than standard chat queries and is useful for everything from professional analysis to creating detailed travel or gift guides.
- Use Voice Mode’s Full Potential: Don’t just talk to the AI. Use the powerful multimodal capabilities of voice mode on the mobile apps (especially Gemini and ChatGPT) by pointing your camera or sharing your screen to get real-time help with visual tasks, from solving math problems to identifying objects.
- Prompting Strategy: Complex prompt engineering is less important now. Instead, focus on providing clear context (by uploading documents), being specific about your desired output, and engaging in a back-and-forth conversation. Don’t be afraid to ask for many variations and use the “branching” feature to explore different conversational paths.
- Generative Capabilities: ChatGPT and Gemini can create images and video (Gemini’s Veo 3 is noted as impressive). All three can produce code, documents, and other complex outputs, sometimes requiring a specific mode like “Canvas.”
Conclusion & Takeaways
To truly benefit from AI, you must move beyond the default settings and casual use. The difference between a casual user and a power user lies in knowing that advanced features exist and actively applying them to real work. The key is to subscribe to a top-tier service, consciously select the most powerful model for important tasks, and integrate features like Deep Research and multimodal voice commands into your workflow. While AIs still have limitations like hallucination, their capabilities for high-level work are immense when used correctly.
Mentoring Question
The author argues that most people use AI like a simple search engine. How does your current usage compare to this, and which specific strategy from the guide (e.g., using a powerful model, trying Deep Research, or using multimodal voice) could you adopt this week to start using AI more like an interactive “thought partner”?
Source: https://www.oneusefulthing.org/p/using-ai-right-now-a-quick-guide?utm_campaign=post&utm_medium=web