AI Assistant — Personal Analyst, Not a Crystal Ball

Imagine having an analyst by your side who can read all five pillars of a stock in seconds, knows what institutional players are doing, understands the market regime, and what the news is saying — and can summarize it in plain English. That's exactly what the AI Assistant / Brain tab is for.

And right at the start, the most important thing:

★ Insight: The AI Assistant does not predict the future. It helps you process and understand existing data. It's a tool for reading and connecting, not a crystal ball. Anyone who treats it like a fortune teller will sooner or later deceive themselves.

---

What the Tab Contains

1) Research Assistant (chat) — 5 queries per day for all plans

You type a query about a stock/situation and the AI responds clearly. Under the hood, it pulls context from QMA data — the five pillars, smart money, market regime, macro, news — and weaves a summary. Instead of clicking through five tabs, you ask and get it all connected.

The limit of 5 queries per day keeps costs under control and forces you to ask quality questions.

2) Brain / Intelligence — under the model's hood

• ML predictions (with probabilities, not certainties).
• Adaptive weights — how much the model trusts each pillar (changes over time).
• Pillar accuracy — how each has historically performed.
• IC (Information Coefficient) — a key number (see below).
• Behavioral insights + market regime.

★ Insight: The Research Assistant is the "talking" layer you use daily. The Brain is the "engine" — showing how much you can currently trust that engine.

---

What It's Good For — and What It's NOT

It Does Well

• Summarizing context (a connected picture instead of five numbers).
• Explaining terms (what does low EQS mean?).
• Connecting signals (does momentum align with institutional behavior?).
• Saving time.

It Does NOT (and Never Will)

• Doesn't know tomorrow. No model knows what the price will do tomorrow.
• Can be wrong.
• It's not a command. The decision is always yours.

★ Insight: Think of AI as a researcher, not a prophet. A researcher brings and organizes materials; what you do with them is up to you.

---

What is IC and Why It's Gold

IC (Information Coefficient) measures how well a given pillar has historically predicted future price movement. We take the pillar's score and compare it with how the stock then developed (7/14/30 days). High IC = the pillar was accurate; IC near zero = coincidence.

• High IC → give more weight to signals.
• IC near zero → currently noise.
• Negative IC → currently works in reverse.

★ Insight: IC is an honesty detector. Instead of "trust all pillars equally," QMA says "this one is hitting now, this one isn't." When AI tells you "great Momentum," but in the Brain, Momentum has a low IC, you know not to overestimate that signal.

---

Why ML Needs Time to Mature

The ML brain learns from real trades — records a prediction, waits for the result, adjusts weights. Short-term signals mature over weeks, long-term over months. When part of the model is fresh, don't take its confidence as a done deal.

★ Insight: ML that claims to be perfect right away is lying. An honest model admits it's learning — and QMA shows you that maturation process.

---

How to Formulate a Good Query (and Not Waste 5)

• Be specific — not "what do you think about the market?", but "how does the combination of momentum and smart money look for [ticker]?".
• Ask about connections, not predictions — not "will it rise?", but "which pillars are driving the score up and which down?".
• Request explanations, not verdicts.
• Combine multiple things into one query.

★ Insight: The quality of the answer depends on the quality of the query. Five targeted queries > fifty random ones.

---

How to Connect — AI ↔ Stock Detail ↔ Tabs

1. Start with the stock detail (raw numbers).
2. Have AI summarize the context.
3. Verify in the Brain — IC of the pillars the answer relates to.
4. Check the relevant tabs (Smart Money, earnings calendar).
5. Decide for yourself.

★ Insight: The strength of QMA is not in one magic number, but in the fact that three perspectives meet — understandable summary (AI), reliability measure (Brain/IC), and raw detail (tabs).

---

Tips

1. Let AI do the boring work, not the important decision.
2. Read IC before trusting a signal.
3. Respect the model's maturation.
4. Cross-verify.
5. Save queries — 5 well-thought-out > 50 random.

---

Common Mistakes

Blind faith in AI · taking output as certainty · not verifying it · asking for predictions ("will it rise?") · ignoring contradictory signals (that's valuable information "be cautious").

---

Summary

The AI Assistant is a personal analyst that can read platform data and summarize it clearly. The Brain shows how much you can currently trust that model (IC, pillar accuracy, ML maturation). Use both as tools for understanding and connecting, not as fortune tellers. Ask specifically, cross-verify, respect the limits — and always keep the final decision to yourself.

---

Disclaimer: The AI Assistant and Brain are not investment advisors. Outputs are analytical and educational, may contain errors, and describe past or current data — not future developments. Always verify information from multiple sources, and any investment decision and its risks are yours. QMA is an analytical and educational tool, not a recommendation to buy or sell.