How to Think Clearly In The Era Of AI: Full Course (5 Hours)

Key Wellness + Cognitive Clarity Strategies (from the course)

Core idea: “Thinking clearly” depends on 3 cognitive functions

• Attention — the gatekeeper/spotlight for what enters awareness
• Working memory — limited “workbench” capacity; improved via meaning and “chunking”
• Executive function — the foreman/manager that decides what to keep/discard and when to switch

---

Hardware: Health Foundations That Improve Cognition

Fix sleep first

• Rule out sleep-disordered breathing
  • Signs include: snoring, waking tired, mouth breathing, narrow jaw, and lots of caffeine
• Aim for 7–9 hours
• Use a consistent wake time
  • Align with your circadian clock
• Stop caffeine ~10 hours before bed
  • Example: stop after noon if you sleep around 10pm
• Bright mornings + dim evenings
  • Get natural light soon after waking
  • Reduce stimulating light at night
• Keep the bedroom cold, dark, and quiet
• Consider bloodwork for nutrient deficiencies and metabolic factors
  • Examples: iron/ferritin, B12, vitamin D, thyroid, and blood sugar markers
• Reduce alcohol
  • Disrupts sleep quality, especially REM
• Prioritize protein + fiber
  • Protein: stabilizes blood glucose and increases fullness
  • Fiber: supports gut microbiome and mood/cognition
• Supplements (only if evidence-backed / low-risk)
  • Creatine: about 3–5 g/day
    • Often cited for a ~1–3% cognitive benefit (more noticeable when sleep-deprived)

---

Exercise as a cognitive amplifier

• Track/optimize VO₂ max
  • Cardio fitness proxy for brain oxygen-metabolism capacity
• Add REHIT (reduced exertion HIIT/sprint intervals)
  • Short hard bursts (e.g., ~10–20 sec sprints) with easy recovery
  • Done a few times per week; presented as high ROI for VO₂ max improvement
• Stack defaults
  • Combine habits so one action improves multiple constraints
  • Example: morning outdoor exercise boosts light exposure and circadian timing

---

Workspace & Environment Design (Low Effort, High ROI)

Improve air quality

• Use a CO₂ monitor and ventilate if CO₂ rises too high
• Or use a ventilation routine (e.g., crack a window)

Reduce intelligible speech during deep work

• Avoid talking, podcasts, or music-with-lyrics during deep focus
• Use masking (brown noise, fans, rain) or physical barriers (closed door/earplugs)

Get light exposure at work

• Place your desk near a window
• Use bright artificial light if needed

Reduce physical + visual clutter

• Clutter competes for attention and burdens working memory

Batch notifications (avoid attentional residue)

• Turn off banners/sounds
• Use scheduled summaries or check phones at set times (e.g., 2–3 batches/day)

Single-task your digital environment

• Full screen; close extra tabs/apps
• Reduce on-screen “decision friction”
  • Email/chat constantly available increases context switching costs

---

Behavioral Strategy: Willpower Isn’t the Main Lever—Opportunity and Friction Are

Instead of treating willpower as a depleting reservoir, treat it as a symptom. The real lever is reducing opportunities to make mistakes via environment engineering.

Willpower replacement: Choice architecture + friction

• Choice architecture / defaults
  • Distractions exist because systems were designed to maximize someone else’s KPI (e.g., watch time, clicks, ad exposure)
  • You can redesign your defaults: tabs, phone settings, app permissions, calendar structure
• Friction
  • Add friction to unwanted behaviors
  • Remove friction to desired behaviors
  • Make the desired action easier once (through engineering), not via constant self-control
• Ulysses (Odysseus) contracts
  • Pre-commit earlier (when strong) to restrict options later (when tempted)
  • Examples:
    • Put your phone in another room
    • Use accountability systems (announced publishing/work schedules, social commitments)
    • Use financial stakes (prepay/donate to a party you dislike if you fail)

---

Key Productivity & Decision-Making “Software”

Working memory improvement via meaning (“chunking”)

• Convert raw data into meaningful chunks
  • Example: letters → words → concepts
• High performers remember more within their domain by chunking at higher abstraction levels

---

Expected Value (EV) for better decisions than “gut”

• Compute:
  • EV = Σ (probability × outcome value)
• Purpose:
  • Replace untracked gut decisions with falsifiable, revisable predictions
• Warning:
  • Avoid resulting (judging decisions only by outcomes on small samples)

---

Power law + chain law for leverage

• Power law: most results come from a small “head” of drivers
  • Double down on top levers
• Chain law: reliability depends on the weakest link
  • Harden bottlenecks; reduce the number of steps

---

Local vs global maxima

• You can be “successful” but stuck on a smaller peak
• Signs you’re near a local max:
  • Diminishing returns
  • More effort but not more output
  • Others with less skill outperform you (different hill)
• Strategy to reach higher peaks:
  • Step down safely (quantify the temporary downside)
  • Transition in parallel
    • Example: dedicate 10–20% time exploring while maintaining the current engine

---

Attentional residue (reduce task-switch tax)

• After switching tasks, residue remains and harms the next task (especially with frequent interruptions)
• Fix:
  • Write a resume plan before switching (next step when you return)
  • Limit open tasks (rule: no more than two open tasks)
  • Avoid context switching when possible

---

Acrasia → “if-then” automation

• Acrasia: knowing what you should do but not doing it
• Fix:
  • Make next actions tiny, specific, and triggered automatically
  • Taps / Trigger Action Plans (if X, then Y)

---

Actionable Techniques Explicitly Recommended (Often with Step Formats)

Taps (Trigger Action Plans): implement as

• When (trigger) happens:
• I will (action) immediately do the smallest physical step

Trigger design criteria

• Obvious (can’t miss)
• Visceral (sensory cue)
• Reliable (fires consistently in the right context)
• Small + physical (start, don’t “solve”)

Installation method

• Rehearse visually ~10 times
• Optionally practice in the real environment
• Add taps in chains (one tap cues the next)

---

Noticing (confusion/flinch/should)

• Label immediately:
  • “I notice I’m confused”
  • “I flinched”
  • “I said should”
• Then ask:
  • Why is it happening?
  • What concrete step follows?
• Purpose:
  • Align your “map” to territory and improve over time

---

Premortems (planning by imagining failure)

• Pretend it’s later and the project failed:
  • “It failed—why?”
• Generate failure stories, categorize causes, then apply fixes
• Use AI as a red-team generator for failure modes

---

Resolve cycles (5-minute timers)

• When stuck:
  • Set a 5-minute timer
  • Attempt progress immediately
  • If progress isn’t possible, create the next 5 steps as bullets
• Goal:
  • Destroy “I’ll get back to it” by forcing a clear next action

---

Red-teaming with AI (without outsourcing final decisions)

• Don’t ask: “Is this good?”
• Ask adversarial questions such as:
  • “Assume this plan is a mistake—top reasons not to do it?”
  • “Out of 100 similar attempts, how many succeed?”
  • “Pretend it failed 6 months later—write the postmortem.”
• Keep responsibility:
  • Write reasoning in your own words
  • Stay accountable for consequences

---

Presenters / Sources Mentioned

• Nick (spoken as “Nick”): business owner and course author; behavioral neuroscience background; references “Claude” / “Claude Fable” for diagrams
• Sophie Loi: attributed for coining “attentional residue”
• Peter Gollwitzer: implementation intentions / if-then plans
• Annie Duke: term “resulting”; referenced book Thinking in Bets
• Scott Alexander: LessWrong; references a quote about inconvenience
• Frederick Bastiat: opportunity cost / “broken windows” story (and “what is seen is seen…” quote)
• Charles Goodhart / Goodhart’s law: “When a measure becomes a target…”
• Pareto: Pareto principle / 80–20 concept
• Chesterton: Chesterton’s fence parable
• Ulysses / Odysseus: Odysseus contract metaphor via The Odyssey
• Aristotle: acrazia discussed in Nicomachean Ethics
• Allen: CO₂/air-performance study attribution in the transcript
• 1997 enrichment study (mice)
• Additional measurement references: VO₂ max; Cooper equation (mentioned as an estimate formula)