How to Use AI to Think and Decide Better

TL;DR

Use AI to improve the information you consume, challenge your assumptions, and strengthen decisions, rather than merely completing routine tasks faster. Curate higher-quality inputs, create focused daily briefings, learn topics through NotebookLM, and red-team important plans by identifying fatal flaws, exposing blind spots, ranking risks, and preparing contingencies.

Transcript

Most people think they're using AI the smart way, but they're actually getting dumber. The smartest people I know aren't using AI to get answers. They're using it to upgrade their entire brain. And the gap between them and everyone else is growing every single day. I've built AI into every part of how I run my business. I'm talking my team, my lear... Read More

Key Insights

• AI is more valuable as a thinking partner than as a task calculator. Using it only for emails, summaries, or slide decks captures surface-level efficiency, while using it to improve learning, reasoning, and decisions can strengthen the person directing the work.
• The quality of ideas is directly connected to the quality of informational inputs. Doom scrolling and random content encourage weak ideas, while frameworks, expert knowledge, targeted news, and relevant insights give the mind better material for generating useful conclusions.
• A social feed is an AI-powered reflection of prior behavior. Resetting suggested content, searching for subjects worth mastering, and deliberately liking, saving, and commenting on relevant material can retrain the platform to deliver information that better supports learning goals.
• A daily AI briefing can concentrate important developments into a manageable routine. The suggested prompt requests the top three developments in chosen categories, two-sentence summaries, source links, an explanation of why each matters, and a format readable in under three minutes.
• NotebookLM supports accelerated consumption by creating a focused AI resource around supplied topics and materials. It can answer questions and convert research into slides, infographics, quizzes, flashcards, or a podcast, allowing information to be consumed in a format suited to the learner.
• Just-in-time learning focuses research on an immediate decision instead of collecting information for hypothetical future use. The approach is designed to provide relevant knowledge when it can be applied that same day, making consumption more purposeful and closely connected to action.
• AI red teaming works by intentionally attacking a plan before external competitors or circumstances expose its weaknesses. A premortem asks why the project failed six months later, encouraging backward analysis from the imagined disaster to the vulnerable assumptions or single points of failure.
• The human role in AI-assisted work is described as director rather than doer. The transcript assigns people responsibility for taste, vision, and care, including judging quality, imagining a future that does not yet exist, building emotional connection, and enrolling others in an idea.

Questions & Answers

Q: How can AI improve thinking instead of just saving time?

AI can improve thinking by upgrading both inputs and outputs. For inputs, it can collect relevant developments, summarize expert material, and support focused learning. For outputs, it can challenge assumptions, simulate a competitor, expose weaknesses, and rank risks. This turns AI from a tool for completing routine tasks into a partner for learning, reasoning, and decision-making.

Q: How do you create a daily AI research briefing?

Prompt AI to act as a daily research assistant and identify the top three developments in the categories most important to you. Ask for each development to be summarized in two sentences with source links, followed by an explanation of why it matters. Request a clear, entertaining briefing that can be read in under three minutes each morning.

Q: How can you reset social media feeds for better learning?

Open the platform’s content preferences and reset its suggested content when that option is available. Then search for subjects you genuinely want to master and deliberately engage with useful posts by liking, saving, and commenting. These actions tell the recommendation system what you value, gradually turning the feed into a more focused source of educational inputs.

Q: How does NotebookLM support faster, focused learning?

NotebookLM can organize supplied information around a particular topic into a focused resource that users can question conversationally. According to the transcript, it can also transform the research into PowerPoint slides, an infographic, a quiz, flashcards, or a podcast. This supports accelerated consumption by presenting the same knowledge through formats that match a person’s needs.

Q: What is the difference between just-in-time and just-in-case learning?

Just-in-case learning means consuming books, conversations, or other information because it might become useful someday. Just-in-time learning begins with a current need, such as a decision that must be made later that day, and gathers information specifically for that purpose. The second approach connects learning to immediate application instead of indefinite accumulation.

Q: How do you red-team a business plan with AI?

Start with a premortem by asking why the project failed six months in the future. Next, instruct AI to act as a cynical, highly successful competitor and explain how it would exploit weaknesses to steal customers. Finally, ask it to rank the top three risks by likelihood and impact, then create a contingency plan for each one.

Q: Why should obvious decisions receive additional scrutiny?

Decisions that feel obvious may be shaped by hidden assumptions, emotional attachment, or blind spots. The transcript argues that people naturally notice evidence supporting their own ideas while overlooking fatal flaws. AI can slow this process down by questioning the reasoning, identifying failure scenarios, and forcing the decision-maker to explain why a conclusion appeared correct so quickly.

Q: What should humans contribute when AI handles routine work?

Humans should operate as directors who provide taste, vision, and care. Taste means recognizing what looks or feels excellent. Vision means imagining a desirable future that does not yet exist. Care includes emotional connection, communicating with people, presenting ideas, encouraging involvement, and enrolling others. These contributions guide and refine the work AI produces.

Summary & Key Takeaways

• AI provides greater value when it improves thinking instead of merely accelerating emails, summaries, presentations, and similar tasks. Better results begin with premium inputs, including expert knowledge, useful frameworks, and relevant industry developments. The central principle is that the quality of a person’s ideas reflects the quality of information being consumed.

• Three practices can improve informational inputs: resetting recommendation algorithms around subjects worth mastering, creating a short AI-generated daily research briefing, and using NotebookLM to transform focused source material into interactive learning formats. These methods replace random consumption with timely knowledge selected for current goals, questions, and decisions that require attention.

• Important plans should be red-teamed before implementation. AI can conduct a premortem, adopt the perspective of a successful competitor, identify exploitable weaknesses, and rank risks by likelihood and impact. The human role then shifts toward directing AI while contributing taste, vision, care, emotional connection, and the ability to involve other people.