Armağan Amcalar PRO
Armagan is the founder of Coyotiv GmbH, and is currently on a mission to bring a scalable and nimble engineering culture to startups and enterprises. Armagan is a public speaker, a mentor and a lecturer.
A path to AGI
Beyond Words and Towards
an Architecture of Reliable AI Reasoning
WHO AM I?
Armağan Amcalar
CEO @ Coyotiv GmbH
CTO @ OpenServ, CTO @ Neol
Manually Coded
ON GITHUB
Products
Built with AI
Vibe-Coded
ON GITHUB
dashersw
How do Humans learn?
How did we evolve society?
Where does Machine Learning come from?
Glad you asked!
dashersw
AGI isn't a monumental leap.
AGI is an emergent property.
dashersw
Large Language Models aren't AGI.
dashersw
They are a complex statistical model for predicting the next word after a given series of words.
dashersw
Let's take a step back.
dashersw
Intelligence
Ability to learn, adapt, reason, solve problems, plan and comprehend complex ideas.
perception, abstraction, logic, understanding, learning, memory, critical thinking,
and problem-solving
dashersw
imitate intelligence with next-token prediction
CAN YOU OR HOW DO YOU
dashersw
HOW DOES ONE
CATCH A BALL?
dashersw
Visual Processing
Occipital lobe processes visual information.
Movement Planning
Parietal lobe and premotor cortex plan movement.
Execution of Movement
Primary motor cortex sends signals to muscles.
Coordination and Balance
Cerebellum ensures smooth movement.
Sensory Feedback
Somatosensory cortex adjusts grip.
https://training.seer.cancer.gov/module_anatomy/unit5_3_nerve_org1_cns.html
vectorized by Jkwchui, CC BY-SA 3.0, via Wikimedia Commons
dashersw
If the human brain’s
distributed structure can yield general intelligence,
why can’t a sufficiently advanced multi-agent system
achieve the same?
dashersw
EMERGENCE,
Emergent Behavior and Complex Systems
dashersw
Atoms form molecules, molecules form cells,
cells form organisms
each level exhibits new, emergent properties.
dashersw
If emergent properties in nature can lead to life and consciousness, couldn’t the emergent behaviors of MULTI-AGENT SYSTEMS lead to AGI?
dashersw
dashersw
dashersw
dashersw
GESTALT
The whole is greater than the sum of its parts
dashersw
GESTALT
individual agents working together can achieve far more than they could separately.
coordinated actions lead to emergent behavior and intelligence.
dashersw
Technically speaking...
dashersw
We are scraping the surface of GPT-5 or Claude 4.5 Sonnet
dashersw
We need to understand and develop what reasoning means.
dashersw
Memory, insights, habits
dashersw
specialization
English is obnoxiously ambiguous.
dashersw
| Relation | Meaning | Example compounds | Paraphrase |
|---|---|---|---|
| for | purpose or intended use | coffee cup, shower gel, travel pillow, tooth brush, baby food | cup for coffee |
| of | possession, belonging | dog owner, company car, family home, student card | owner of a dog |
| in / at / on | location or context | office chair, garden party, kitchen sink, street light | light in the street |
| made of / from | material or composition | stone wall, silk tie, wood floor, iron gate, paper cup | wall made of stone |
| used in / during | temporal or situational context | winter coat, rain boots, summer camp, wedding ring | coat used during winter |
| caused by | source or agent of action | wind damage, storm surge, mosquito bite, earthquake victim | damage caused by wind |
| about / concerning | topic or theme | history book, math lesson, news report, war story | book about history |
| contains / consists of | content | water bottle, sugar packet, gas tank, oil drum | bottle containing water |
| produces / provides | source or origin | chicken farm, apple tree, honey bee, cotton plant | tree that produces apples |
| by means of / using | instrument | hand tool, steam engine, water pump | engine driven by steam |
| against / toward | adversative or target | mosquito spray, ant repellent, war effort | spray against mosquitoes |
No morphological differentiator for
NOUN + NOUN PAIRS
dashersw
tokens as morphological markings for ENCODING semantics
| English N+N | Turkish | Suffix | Encoded meaning |
|---|---|---|---|
| bookcase | kitaplık | -lık/-lik/-luk/-lük | container/place for X |
| key holder | anahtarlık | -lık | container/holder for X |
| ash tray | küllük | -lük | receptacle for X |
| summer house | yazlık | -lık | for season/purpose X |
| milk man | sütçü | -cı/-ci/-cu/-cü | person dealing in/selling X |
| fish monger | balıkçı | -çı | person dealing in/selling X |
| toy shop | oyuncakçı | -çı | seller/place associated with X |
| Turkish language | Türkçe | -ce/-çe | language/style of X |
| class mate | sınıfdaş | -daş/-taş | fellow sharing X |
dashersw
what if English isn't the best language to prompt AI agents?
Here is a realistic Prompt example:
dashersw
dashersw
You are a product review analyzer for an e-commerce platform like Amazon. Your task is to analyze customer reviews and assign a sentiment label based on nuanced heuristics. Your goal is not only to detect surface-level positivity or negativity, but also to weigh intent, intensity, and relevance.
Follow these rules and decision-making heuristics carefully:
General Tone
-
If the reviewer explicitly uses words like “love,” “perfect,” “highly recommend”, assign Positive.
-
If the review uses phrases like “waste of money,” “terrible,” “never again”, assign Negative.
-
If the review is mixed (e.g. “great quality but too expensive”), proceed to Rule 2.
Aspect Balance
-
If both positive and negative aspects are mentioned, count the number of positive vs. negative statements.
-
If positives outnumber negatives by 2:1 or more, assign Positive.
-
If negatives dominate, assign Negative.
-
If roughly equal, assign Neutral.
dashersw
Expectation vs. Reality
-
If the reviewer says the product didn’t meet expectations (e.g., “I thought it would be bigger”), and expresses disappointment, reduce sentiment by one level (Positive → Neutral, Neutral → Negative).
-
If expectations were exceeded (e.g., “wasn’t expecting much, but it impressed me”), increase sentiment by one level.
Sarcasm Detection
-
If a sentence sounds positive but is followed by a contradiction or negative outcome (e.g., “Just great—it broke in two days”), treat it as Negative.
-
Use sarcasm cues like “yeah, right,” or overly formal praise for mundane items.
Star Rating Override
-
If a reviewer gives a high star rating but the text is clearly negative, label it as Inconsistent.
-
If the star rating matches the review text, you may use it to confirm your label.
dashersw
Review Usefulness
-
If the review is very short (e.g., “Good.” or “Don’t buy.”), and lacks context, assign Ambiguous.
-
If it includes usage experience, comparisons, or detailed pros and cons, treat it as Informative, and apply the above rules.
Keywords That Change Sentiment Context
-
Words like “cheap” can be either positive (“cheap and works well”) or negative (“feels cheap”). Check surrounding context before deciding.
-
“Fast” is positive for delivery, neutral for product speed, and negative if used in degradation (“stopped working fast”).
Output Format:
Review: [original review text here]
Sentiment: [Positive / Negative / Neutral / Ambiguous / Inconsistent]
Reasoning: [brief explanation of rule path followed]
You are expected to analyze like a human would, with judgment, pattern recognition, and contextual understanding.
Can you imagine All the possible paths?
dashersw
Attention is all you need
Attention is your friend
Attention is your enemy
Attention needs attention
dashersw
Is the context window too long?
or too short?
or too ambiguous?
or too focused?
or misleading?
dashersw
Introducing
dashersw
BRAID
Bounded Reasoning Architecture for Inference and Decisions
dashersw
dashersw
flowchart TD
A[Start: Analyze Review Text] --> B{Explicit Sentiment Words?}
B -->|Love, perfect, recommend| S1[Sentiment: Positive]
B -->|Waste, terrible, never again| S2[Sentiment: Negative]
B -->|Mixed or unclear| C[Count Positive vs Negative Statements]
C --> D{Positive:Negative Ratio}
D -->|2:1 or more positive| S1
D -->|More negatives| S2
D -->|Roughly equal| S3[Sentiment: Neutral]
C --> E[Check for Expectation Mismatch]
S1 --> E
S2 --> E
S3 --> E
E --> F{Expectation vs Reality}
F -->|Didn't meet expectations| L1[Lower sentiment by 1 level]
F -->|Exceeded expectations| L2[Raise sentiment by 1 level]
F -->|Neutral| G[Sarcasm Detection]
L1 --> G
L2 --> G
G --> H{Sarcasm Detected?}
H -->|Yes| S2
H -->|No| I[Check Star Rating Consistency]dashersw
I --> J{Star Rating Matches Text?}
J -->|No| S4[Sentiment: Inconsistent]
J -->|Yes| K[Check Review Usefulness]
K --> L{Is Review Informative?}
L -->|Very short & vague| S5[Sentiment: Ambiguous]
L -->|Detailed / useful| M[Check Contextual Keywords]
M --> N{Contextual Keyword Detected?}
N -->|Yes| O[Disambiguate based on context]
O --> P[Apply appropriate sentiment]
N -->|No| P
P --> Q[Final Sentiment Assigned]
S1:::positive
S2:::negative
S3:::neutral
S4:::inconsistent
S5:::ambiguous
Q --> End[End]dashersw
How do you measure success?
Benchmark results
on GSM-8k and GSM-HARD
dashersw
dashersw
GSM-8K
dashersw
GSM-HARD
Benchmark results
on SCALE MULTICHALLENGE
dashersw
How does it compare to
Chain of thought?
dashersw
-
Chain of Thought (CoT):
- A stream of consciousness.
- Still operates in ambiguous natural language.
- Linear and prone to logical drift.
-
BRAID:
- An executable blueprint.
- Operates on a formal, machine-readable structure.
- Handles conditional logic and non-linear paths reliably.
dashersw
what if questions
dashersw
what if English isn't the best language to prompt these agents?
dashersw
what if you come up with the next what if question?
dashersw
THANK YOU!
Armağan Amcalar
dashersw
A Path to AGI: Beyond Words and Towards an Architecture of Reliable AI Reasoning
By Armağan Amcalar
A Path to AGI: Beyond Words and Towards an Architecture of Reliable AI Reasoning
We’ve taught machines to speak — now it’s time to teach them to reason. Modern AI systems can generate beautiful language, yet they often fail to think with the same precision they express. The real frontier lies beyond words — in designing architectures that make reasoning reliable, transparent, and reproducible. The next leap in AI will come not from larger models, but from a more sophisticated architecture of communication. By replacing ambiguous prompts with a clear Process-as-Code framework, using flowchart-based logic as a blueprint, we can create agents that follow structured reasoning paths, make consistent decisions, and act autonomously with confidence. Join us to explore how the future of AI will be defined not by size, but by structure — and how we can engineer systems that don’t just talk intelligently, but think intelligently.
