There is something almost ironic about modern trading.

We have more computing power, more data, faster networks and more sophisticated AI than ever before...

Yet millions of traders are still looking at markets through essentially the same interface:

Candlesticks moving from left to right.

That is exactly where my latest video starts.

A completely naked MetaTrader 5 environment.

One chart.

One-minute candles.

Gold Futures.

Nothing else.

And then, progressively, I add the layers that have shaped the evolution of my own work:

Price → DOM → Order Flow → Market Microstructure → 3D → AI → Auto-Learning

The result is almost a visual timeline of how my approach to market analysis has changed.

🎥 New video:
https://youtu.be/IT35npf04PI

Stage 1: The Market as Most Traders Still See It

The starting point is intentionally primitive.

A standard MT5 candlestick chart.

No indicator soup.

No complicated dashboard.

No 3D.

No AI.

Just price.

And there is nothing inherently wrong with that.

Candlesticks remain one of the most efficient abstractions ever created for financial markets.

But they are still an abstraction.

A candle tells us what price did during a period.

It does not directly reveal everything that happened underneath it.

A single one-minute candle may contain thousands of individual events:

• Market orders

• Limit orders

• Order cancellations

• Liquidity replenishment

• Bid/Ask imbalance

• Pulling

• Stacking

• Absorption

• Icebergs

• Aggressive executions

• Rapid changes in market depth

All of this activity is compressed into four numbers:

Open. High. Low. Close.

That compression is extremely useful.

But it also destroys information.

And that is where my work started moving in another direction.

Stage 2: Exposing the Order Book

The second part of the video introduces my NEXUS_DOM Heatmap for MT5.

This is where the market suddenly stops looking two-dimensional.

Not visually yet, but informationally.

Instead of observing only where price traveled, we begin looking at the liquidity environment through which price is moving.

The chart becomes surrounded by information about the actual microstructure.

You begin seeing things such as:

Bid liquidity

Ask liquidity

Depth concentration

Absorption zones

Liquidity stacking

Liquidity pulling

Volume distribution

Order Flow

Trade activity

Potential iceberg behavior

Book imbalance

Suddenly the question is no longer simply:

“Is price going up or down?”

The question becomes:

“What is happening inside the order book while price is moving?”

That is a radically different way of looking at markets.

Price Is the Output. Microstructure Is the Machinery.

Imagine looking at a car moving down a road.

A candlestick chart shows you:

position + speed + direction.

Order flow begins opening the hood.

You can start observing the machinery generating the movement.

In markets, that machinery is constantly changing.

Liquidity appears.

Liquidity disappears.

Orders get executed.

Orders get canceled.

Large passive participants absorb aggressive flows.

Other participants sweep multiple price levels.

The balance between resting liquidity and aggressive execution continuously changes.

And price emerges from that interaction.

That is why I became increasingly interested in Market Depth and Order Flow rather than simply creating more traditional indicators.

Stage 3: Giving Market Microstructure a Third Dimension

Then comes the third stage.

3D_NEXUS_META.

This was not originally about making charts look futuristic.

The underlying question was much more practical:

What happens if market information is no longer forced into a flat chart?

Traditional platforms have an enormous constraint.

Everything must fit into X and Y.

Time.

Price.

And then we keep adding more panels underneath:

Volume.

Delta.

CVD.

DOM.

Footprint.

Heatmaps.

Signals.

Statistics.

At some point, the screen becomes a cockpit designed by an octopus.

So I started exploring another idea.

What if depth itself became spatial?

What if liquidity could have height?

What if volume became geometry?

What if Bid and Ask activity formed landscapes?

What if individual transactions could exist as objects inside a navigable environment?

This is where 3D_NEXUS_META emerged.

🌐 https://metaquantuniverse.com/nexus

Instead of another chart overlay, the objective became to construct a multidimensional representation of market microstructure.

Why 3D Matters

3D is often dismissed as cosmetic.

And if you simply transform a line chart into a rotating object, it probably is.

But multidimensional visualization becomes interesting when the additional dimension actually represents additional information.

For example:

X axis → Time

Y axis → Price

Z axis → Liquidity / Volume / Intensity / Depth

Now information that previously had to be represented through colors, separate panels or tiny labels can become part of the actual geometry.

The market starts becoming a terrain.

Large liquidity zones can appear as structures.

Trade clusters can become spatial concentrations.

Bid and Ask activity can occupy different visual regions.

Order Flow can become something you explore rather than simply read.

This is closer to scientific visualization than traditional technical analysis.

And this distinction matters.

The goal is not:

“Make trading charts look cool.”

The goal is:

Increase the amount of market structure the human brain can perceive simultaneously.

Then AI Changed the Equation Again

The arrival of modern AI models introduces another major shift.

For decades, trading software has mostly followed this architecture:

DATA ↓ FIXED RULE ↓ INDICATOR ↓ SIGNAL

For example:

RSI < 30 ↓ OVERSOLD

The rules exist before the market data arrives.

The system simply applies them.

Modern machine learning allows a different architecture:

MARKET DATA ↓ FEATURES ↓ PATTERNS ↓ STATISTICAL EVALUATION ↓ MODEL ↓ CONTINUOUS RE-EVALUATION

And increasingly:

OBSERVE ↓ DETECT ↓ MEASURE ↓ REMEMBER ↓ COMPARE ↓ LEARN ↓ ADAPT

That last loop is particularly important.

Auto-Learning: Moving Beyond Static Indicators

Suppose the system detects a particular sequence:

Large Ask Liquidity ↓ Aggressive Buying ↓ Absorption ↓ Liquidity Pull ↓ Delta Shift

A static indicator might simply output:

SELL SIGNAL

But an auto-learning research architecture can ask much more interesting questions.

What happened afterward?

After:

1 second?

3 seconds?

10 seconds?

30 seconds?

60 seconds?

How large was the excursion?

What was the maximum favorable excursion?

What was the maximum adverse excursion?

Did the pattern work during high volatility?

Did it fail during trending regimes?

Was the result stable across sessions?

Was it statistically meaningful?

Does the behavior still exist?

Now repeat that process thousands of times.

That is a fundamentally different paradigm.

The software stops behaving like an indicator.

It begins behaving more like a continuous market research laboratory.

From Trading Platform to Observation Engine

This is probably the most important evolution in my own work.

Originally, the objective was primarily visualization.

Then came Order Flow.

Then Market Depth.

Then quantitative metrics.

Then event detection.

Then 3D.

And now AI increasingly transforms the architecture itself.

The system can progressively become something closer to:

MARKET │ ▼ REAL-TIME DATA │ ┌───────┴────────┐ ▼ ▼ PRICE DOM │ │ └───────┬────────┘ ▼ ORDER FLOW │ ▼ MICROSTRUCTURE │ ┌───────┴─────────┐ ▼ ▼ 3D VISUALIZATION QUANT ENGINE │ ▼ AI MODELS │ ▼ AUTO-LEARNING │ ▼ MARKET RESEARCH

This is where things become much more interesting.

Because the visualization, the quant engine and the AI system are no longer isolated tools.

They can become different interfaces into the same underlying market state.

Humans and AI See Markets Differently

There is another reason I believe this architecture matters.

Humans are extremely good at recognizing certain visual structures.

We instantly notice:

Clusters.

Symmetry.

Outliers.

Concentrations.

Acceleration.

Changes in shape.

AI systems, meanwhile, can continuously measure things humans cannot realistically track manually.

Thousands of events.

Multiple horizons.

Multiple regimes.

Long historical comparisons.

Statistical distributions.

The interesting future is therefore probably not:

Human versus AI.

It is:

Human visual cognition + machine-scale statistical observation.

The 3D interface gives the human a richer representation.

The AI system gives the machine a continuous analytical layer.

Those two systems can complement one another.

Why Gold Futures?

The new video was recorded on the 100 oz Gold Futures market traded through CME / COMEX.

Gold is particularly interesting for this type of demonstration because its microstructure can shift rapidly.

You can observe:

Sudden liquidity concentrations.

Fast book changes.

Aggressive bursts.

Absorption.

Liquidity migration.

Depth imbalance.

Rapid responses around important price levels.

A simple candle may look almost trivial.

But underneath it, the order book can resemble a tiny electronic battlefield operating at machine speed.

That contrast is precisely what the video demonstrates.

Three Screens. Three Generations.

The entire evolution can ultimately be summarized in three images.

1. MT5

PRICE

You observe what happened.

2. NEXUS_DOM

PRICE + LIQUIDITY + DEPTH + ORDER FLOW

You begin observing what is producing the movement.

3. 3D_NEXUS_META

PRICE + ORDER FLOW + MARKET DEPTH + MICROSTRUCTURE + 3D + QUANT + AI + AUTO-LEARNING

You begin constructing an environment capable of studying the market itself.

That is a very different destination from where the project started.

The Bigger Picture

Financial markets are increasingly electronic.

Increasingly algorithmic.

Increasingly driven by machines interacting with other machines.

Execution takes place at time scales far below human perception.

Yet our primary interface with this world is still often a charting paradigm inherited from paper markets.

There is an interesting mismatch here.

Markets evolved.

The visualization paradigm barely did.

I believe the next generation of trading technology will increasingly combine:

Real-time Order Flow

Full Market Depth

High-frequency event detection

Multidimensional visualization

Local AI models

Specialized micro-models

Continuous quantitative research

Auto-learning systems

And perhaps most importantly:

systems that can challenge their own assumptions.

Not simply:

“Here is my signal.”

But:

“I detected this pattern 1,284 times. Here is what happened afterward, here is its distribution, here is how it behaved across regimes, and here is whether the edge appears to be degrading.”

That is much closer to how I imagine the future of trading research.

And This Is Still Only the Beginning

This new video is therefore more than another demonstration of 3D_NEXUS_META.

It captures a transition in the philosophy behind the project.

From:

Charting

to

Visualization

to

Microstructure

to

Quantitative research

to

Artificial Intelligence

And ultimately toward systems capable of continuously learning from the markets they observe.

The chart is slowly turning into something else.

A sensor.

A laboratory.

A visualization engine.

An AI interface.

And perhaps eventually...

a real-time model of the market itself.

🎥 WATCH THE NEW VIDEO

I Turned MetaTrader 5 Into a 3D AI Trading System

▶️ https://youtu.be/IT35npf04PI

🌐 DISCOVER 3D_NEXUS_META

https://metaquantuniverse.com/nexus

From candles to liquidity.
From liquidity to microstructure.
From microstructure to 3D.
From 3D to AI.

That is the direction. ⚡🧠🌐