The G Loop: The Universal Algorithm of IQ
Using Confidence (η) and Uncertainty (χ) to Supercharge Problem-Solving and Increase IQ
Cortical Columns
Cortical columns are often described as the brain’s fundamental “microprocessors,” each comprising a vertically arranged ensemble of neurons that together encode, transform and relay information in a remarkably uniform, repeating pattern across the entire cortex. Decades of anatomical, physiological and modelling studies have shown that these minicolumns — think bundles of roughly a few hundred neurons spanning cortical layers — perform the same core computations whether you’re recognising a face in the fusiform gyrus or planning a movement in the motor strip.
The G Loop
At its core, the G-Loop proposes that every cortical column doesn’t simply process one‐off inputs, but instead runs a continuous, 7‐step cycle of sensing, error‐checking, mode‐selection, execution, reflection, adaptation and reset — effectively a universal problem-solving engine. This recursive loop not only steers local microcircuit computations (from sensory encoding through motor planning) but also scales seamlessly across networks of columns, whole brain functional networks, and even between interacting brains, uniting fluid and crystallised intelligence into a single, self-tuning framework. In the diagram below, each node maps onto one phase of this cycle, with “Evolve & Grow” representing the gradual expansion of cognitive capacity that emerges as the loop iterates.
Meta‑Regulators & Brain Rhythm Coupling
Each loop is shaped by two regulators and two intertwined rhythms:
η (Eta = Confidence): Builds with successful results tilting the system toward template-driven execution — whether in routine autopilot or capitalising on unexpected wins.
χ (Chi = Uncertainty): Grows when existing templates have inconsistent clarity, steering the system toward generative mapping — whether to fix emergent problems or explore new strategies.
θ–γ (theta-gamma) brainwave coupling for short term, limited capacity working memory: Supports rapid sequencing of schemas and templates; can underlie both Control Mode (apply or exploiting known strategies) and Creative Mode (quick prototyping of novel ideas).
δ–γ (delta-gamma) brainwave coupling for long-term working memory: Enables deep relational mapping and analogy building; can underlie both Explore (rich generativity) and Exploit (integrating new insights into existing schemas).
Introspective Tip: After phases in the G loop, rate on a 1–5 scale your confidence (η), and your uncertainty (χ). These scores guide decision threshold adjustments on your next cycle.
The G Loop In Action
1. Sense & Move
Define your Problem Space
Imagine you are preparing for a presentation at work. You jot down your goal (“convince the team to adopt a new project management tool”), list constraints (30-minute slot, mixed seniority), note resources (slides, demo account), and recall any relevant history (last tool rollout faltered because training was rushed).
At home, perhaps you're planning a family weekend trip. You mark your destination, budget, child-friendly activities, and lessons from past trips (“avoid too many long drives”).
Why it matters
By capturing facts, feelings, and past attempts, you prime your mind with everything you need. It’s like clearing the start line so your brain can sprint ahead.
η‑Check: Quickly self‑rate your confidence in the plan so far (1–5). This initialises your η.
2. Prediction-Error (Free Energy) Check
Assess the gap
Next, take a quick mental temperature check: are you cruising comfortably, or is something out of reach or ‘off’? Is there a problem to solve ?
Autopilot: If things are smoother than expected (you’ve rehearsed your presentation well, or the weather forecast for your trip looks perfect), you stay on routine.
Fluid Branch: If you hit a snag (your demo account isn’t working, or there’s a thunderstorm warning), you open your mind to more flexible problem-solving.
Example
You arrive early for your talk and find the room’s projector isn’t working! That surprise jolts you out of autopilot into more creative fixes.
χ‑Check: Rate your uncertainty (1 = low, 5 = high). A high χ indicates your δ–γ loop should be engaged.
3. Decision: Autopilot vs Exploit vs Explore
When you encounter a situation requiring active problem-solving (the fluid intelligence forks of the Trident), there are two distinct modes your brain can tap into: Exploit (Control) or Explore (Creative). This decision hinges on whether your initial outcome is unexpectedly good (Exploit) or unexpectedly poor (Explore). If things are running smoothly without surprises, you are in Autopilot, comfortably continuing with well-learned routines - your crystallised intelligence. This is the shaft of the trident.
Here are the branching ‘modes’ you can choose between:
1. Autopilot (Routine & Familiar)
When: Outcomes match your expectations; no surprises or challenges arise.
Action: Keep running your familiar, automatic habits — like giving a rehearsed talk or following your usual recipe for family dinner. Minimal mental effort required.
Example:
At work, delivering a regular weekly briefing that you've done dozens of times without a hitch.
At home, cooking your favourite meal — one you can prepare without even glancing at the recipe.
Exploit (Fluid Mode: Capitalising on Unexpected Success)
When: Outcomes are better than expected but surprising—you notice something working particularly well, beyond your usual experience. This is not routine but rather a promising novelty to build upon quickly.
Action: Rapidly develop and implement a new strategy or insight to harness this unexpected advantage.
Memory engagement:
Long-term memory (LTWM): Quickly retrieve higher-level insights or mental models to understand what's making things go unusually well.
Working memory (STWM): Keep track of concrete, emerging details or steps in real-time to immediately act upon this insight.
Example:
At work, your audience unexpectedly loves a humorous, spontaneous anecdote at the start of your talk—so you swiftly decide to incorporate related humorous examples throughout your presentation.
At home, your children unexpectedly enjoy an educational game you introduced casually — prompting you to quickly develop similar activities to further their engagement.
3. Explore (Fluid Mode: Navigating Unexpected Challenges)
When: Outcomes are worse than expected, and you face unexpected difficulties or barriers.
Action: Open your mind to experimenting, brainstorming, or trying completely new approaches to solve the emerging problem.
Memory engagement:
Long-term memory (LTWM): Construct new analogies or conceptual connections, creatively exploring novel solutions.
Working memory (STWM): Hold in mind temporary hypotheses, notes, or "mini-experiments," testing them as you go.
Example:
At work, your projector completely fails, prompting you to quickly pivot to a whiteboard sketch and interactive Q&A to keep the audience engaged.
At home, rain ruins your outdoor event, triggering you to brainstorm last-minute indoor alternatives, perhaps transforming your planned BBQ into a fun indoor picnic with games.
The Impact of Self-Efficacy η on your Choices
Think of η as your ‘confidence meter’ when you decide to stick with what’s working (Exploit) or try something new (Explore). It’s your gut feeling of “I’ve got this.” It rises when you’ve seen your actions pay off before, and dips when you’ve recently stumbled.
High η (you feel confident)
You need only a tiny hint of success to keep doing what you’re doing. Even if things only improve a little, you’ll go all-in on your current plan — because your confidence threshold is low.Low η (you feel unsure)
You demand a much bigger win before you’ll commit. Otherwise, you’ll keep poking around — trying different approaches — until you feel that confidence bump again.
How it shifts the balance in which ‘mode’ you take:
High confidence (η↑)
Lowers the bar for “Exploit.”
Even small wins feel good enough to stick with, so you jump straight back into the same strategy.
Example: You nailed that recipe last night — so tonight, when your first spoonful tastes just a bit better than usual, you keep cooking exactly the same way.
Low confidence (η↓)
Raises the bar for “Exploit.”
You need a big, clear win before you’ll commit — otherwise you’ll keep experimenting.
Example: You tried a new route to work but got stuck in traffic. Until you find a stretch that really flies, you’ll keep exploring alternative roads each morning.
Why this matters
By tuning whether a win feels ‘good enough,’ self-efficacy makes you more or less adventurous. When you’re confident, you grab onto improvements quickly. When you’re doubtful, you keep your options open until one truly shines.
Think of χ (Chi) as the “inner tug-of-war meter” when different parts of your mind pull you in opposing directions — short-term versus long-term plans, a gut hunch versus cold logic, emotional impressions versus analytic models.
High χ (lots of disagreement)
You notice conflicting signals—your head says one thing, your gut another, and your memory yet a third. That conflict slows you down, raises your “go forward” bar, and makes you dig deeper: you’ll gather more data, compare options, ask for advice, or sketch out side-by-side pros and cons before moving on.Low χ (clear agreement)
All your mental loops line up—your immediate impressions, long-range goals, and rational steps reinforce each other. Decisions flow smoothly: you’re more likely to stay in Autopilot or shift quickly into Exploit without pausing to second-guess.
The Impact of Meta-Uncertainty (χ) on your Choices:
You can think of Chi (χ) as the sense of ‘something isn’t adding up’ when different ways of thinking — short-term vs long-term, emotional vs logical, one mental model vs another — give you mixed signals. χ tunes how big a surprise (positive or negative) has to be before you switch out of Autopilot at all. It raises the decision bar whenever your gut, your goals, and your data aren’t in sync.
How it shapes your choice:
High Chi (χ↑)
Decision threshold ↑: You raise your bar for both Exploit and Explore.
Action: Pause, reflect, compare alternatives, seek fresh perspectives, or gather more info before committing.
Memory engagement:
Long-term working memory (your ‘problem space’ at the time): Revisit broader goals and past lessons.
Short term WM (your limited capacity active memory and attention control): Jot down quick pros/cons, simulate ‘mini-trials’ in your mind.
Example: At work, your gut says scrap the slide deck but your data review says stick with it. You schedule a quick team huddle to weigh both before deciding.
Low Chi (χ↓)
Decision threshold ↓: Coherence lets you switch smoothly — into Autopilot if there’s no surprise, or straight to Exploit/Explore if there is.
Action: Act swiftly on whichever mode the situation demands, with little need for extra checks.
Example: You’ve used the same software process successfully for months; when it flags a tiny improvement, you immediately weave it into your workflow without overthinking.
Why noticing Chi matters
By tuning into that subtle ‘inner static,’ you know when to hit the brakes and get more clarity versus when to trust the flow and move fast. It helps you avoid both rash snap-decisions and endless overthinking — so you stay flexible, focused, and in tune with what matters most. By snagging you whenever your mental loops disagree, Chi helps you avoid both rash gambles and paralysis by analysis. You only dive into a new strategy (Exploit) or tweak your approach (Explore) when your inner voices settle on one clear course.
In summary for these ‘meta-regulators’ in the G loop:
η (self-efficacy) adjusts how big a win you need to exploit.
χ (meta-uncertainty) adjusts how much mismatch between your inner signals you’ll tolerate before you pause to think—raising or lowering the bar for both Exploit and Explore.
Why distinguishing ‘Modes’ matters
Autopilot conserves mental energy by relying on trusted routines.
Exploit capitalises immediately on surprising wins, rapidly locking in beneficial new strategies.
Explore adapts flexibly when facing surprising setbacks, inventing and testing creative alternatives.
Understanding this choice of ‘modes’ helps you swiftly decide how best to respond — whether you're professionally troubleshooting an unexpected opportunity or challenge, or simply adapting to the ever-changing rhythms of daily life.
4. Execute
Now it’s time to execute, harnessing the chosen mode in step 3 above. Particularly for difficult problems, try switching between different representations if useful. These include:
Visual diagrams: Sketch a quick diagram to reveal hidden connections or simplify complexity. For example, drawing the user flow on a whiteboard can help you spot friction points in your presentation.
Tabular grids: Lay out pros and cons in a table to clarify trade-offs — ideal for choosing between two vacation spots.
Narrative analogies: Frame the problem as a short story to reveal causes and effects, helping you empathise with your audience’s concerns.
Mathematical summaries: Simple charts or numbers can show patterns clearly — for instance, graphing previous budget overruns makes financial planning intuitive.
Each of these may run it’s own ‘G loop’. Your intuitions based on autopilot mode may be another G loop that plays out too.
The key is adapting your approach by leveraging multiple representational formats to unlock hidden insights about relationships in the data.
5. Meta-Cognitive Binding
Once you’ve prototyped or applied a solution, your brain asks “How coherent is all this?” That’s χ’s job — the ratio of misaligned to aligned G Loops for different takes on the problem.
Low χ (clear agreement) → your various “inner voices” line up, so you fasten your choice with a winner-take-all “lock,” tucking the new insight into place.
High χ (lots of inner tug-of-war) → you hold off, gather a bit more data or chat with a colleague before you commit.
This binding step uses the very same delta–gamma and theta–gamma rhythms you’ve been juggling — only now they snap together or stay loose, guided by χ against a binding threshold.
6. Consolidation & Meta-Gate Update
With your decision locked in (or postponed), you update your meta-gates for next time:
If you Exploited well (η ticked up) → you ratchet your confidence threshold downward, so small wins tomorrow feel good enough to exploit sooner.
If you Explored richly (χ ticked up) → you widen your uncertainty tolerance, making it easier to entertain big pivots in the future.
These simple updates self-tune your gate knobs so that real-world success and surprises continually refine your decision thresholds.
7. Loop Reset & Next Cycle (Ready for Round 2!)
Finally, the system resets:
Abstract Convergence: η and χ settle toward their new equilibrium.
Pause or Fresh Inputs: Your Loop waits for the next nudge — maybe a new datum, a sudden challenge, or an off-hours “aha!” during a walk.
Start Over: Back to Sense & Move, now armed with sharpened thresholds and enriched rhythms.
And just like that, your brain has run a full micro-algorithm — melding autopilot, rapid locking-in, and creative exploration — while constantly calibrating itself to the edge of surprise and stability in that ‘brain critical’ region we want to be in. Dance on that critical edge, and you stay both reliable and inventive.
With each pass, you’re not just solving today’s problem — you are upgrading your mind for whatever comes next.