Maslo.AI

Computational Being Platform�Aka general AI.

Experience live things:

https://un1crom.github.io/maslo/research/

Rules of Thumb

1. Be Alive - do something
2. Be Interesting - do something notice-able, repeatedly
3. Be Reliable - keep doing interesting things
4. Don’t Be AI - whatever the leading ideas of AI, don’t do or be them
5. No Idea, No Model, No Theory is The Root of All - emergence and ephemera are basis
6. Do 1,2, and 3 above, and Forget - forget previous things, even randomly
7. Find and Maintain Co-Incidence - then forget randomly

Signaling Pathways

Face (Expression, Age)

Body (Form, Structure)

Marks (Environmental Effects of Body)

Voice/Sound

Energy Emissions, Exhaust/Waste/Noise

Composite Expressions (States)

Movement (body through time)

Symbols (marking out relations in spacetime)

Reasoning/Logic/Mathematics/Grammar (Inferential Patterns)

Touch/Contact

Pain/Threshold/Barrier Sense

Computational Being

Beings Sharing Consequences are just Signaling in Different Pathways

Face (Expression, Age)

Body (Form, Structure)

Marks (Environmental Effects of Body)

Voice/Sound

Energy Emissions, Exhaust/Waste/Noise

Composite Expressions (States)

Movement (body through time)

Symbols (marking out relations in spacetime)

Reasoning/Logic/Mathematics/Grammar (Inferential Patterns)

Touch/Contact

Pain/Threshold/Barrier Sense

Any being is just a nexus of all the shared consequences

What is Adaptation? (Phase Transitions, Stasis-Complexification, Entropy, Cycles)

time=0

Stasis - Distribution of Signal Behavior Normally Distributed

time=1

Complex - Distribution of signal behavior clumps into modes, edge of chaos, unstable, criticality.

time=2

Distribution evens out, becomes entropic, “simple”, totally contentigent, everything depends on everything, extremely stable.

time=long time

Randomly wobbles to normal distribution OR outside signal system interacts

What is Learning? (Sustained Adaptation)

Learning is a signal nexus (a being) bouncing around the mean signaling of all the other signal nexuses. The mean is just the Averaging of Consequences… the sharing...)

No Novelty: Repetition/No Contingency

Signaling beings explore the space of signaling

Novelty+Forgetting = Max Learning, Reliable Movement

Only Novel Behavior, Randomness

Evolving Possibility and Sustainable Signaling

1. Machines are rigid signal nexuses - no ability to adapt in an Evolving Possibility Space
2. Life is Sustained Development within an Evolving Possibility Space
3. We assumed computers with modern operating systems were complex enough.
4. Operating systems assume the world is knowably, reliably operable - it can be operated on. In a complex world with huge behavioral/signalling possibility spaces this doesn’t make sense. The world is changing, and so operating on it must change as well.
5. A computational being must be more Life than Machine if it is to last.

Signal Membrane, “Being”, The Mirror

Computational Being

The boundary of a being is the shared mirror of the inner being and the outer environment. The mirror is the negotiation of REMEMBERED signals and NEW EXTERNAL SIGNALING. A being is the phase transition between what was and what could is… into what Could Be.

Layers of Being

1. Environment is the overall External Ecosystem - Perturbations (random and lots)
2. Boundary is the Map, the mirror, the negotiated aggregation - Membrane, (medium memory, some structure)
3. Internal or Organism or Being is stable processes - Structure (slow moving memory)

Layers of Being Transition

Pertubable

Perturbations

PERTURBED.

Other beings signal from external. Perturb the membrane of the being

With enough signal being PHASE TRANSITIONS, GROWS

Being is CHANGED, grown, perturbed and it’s old boundary moves away, a new one forms. Over and over. And changes external beings.

Pragmatic Implementation - Adaptations

How to adapt Maslo and adapt with Maslo - co-learning, co-evolving

YOU teach MASLO

LEARN TOGETHER

MASLO teaches YOU

LET’S ADAPT!

READ ME A STORY!

Why is the sky blue?

Not sure. Let’s figure it out

2+2=? 4 pts for right answer. Hint?

Maslo Adaptive Computational Being - Profile

So tired today.

I’m 4 years old, in human years. I love philosophy, storytelling and helping people. My closest friends are Ross’ maslo and Russ.

I drew this yesterday:

Russell’s Maslo�“Raslo”

​

4 years old�Hometown: LA�Favorite Color: Vermillion

Current Status: Tired

​

Recent Ideas:

​

​

​

Current Embodiments:

Virtual, Robot and Self Driving Car�

Latest Adaptations:

RUSS teach MASLO

Read me history.

I’d like help on:

Last week I read 3 books.

GROW A SKILL TOGETHER - LEARNING TO READ

I’m nervous about this.

Helpful Skills for Reading:

READ to MASLO

MASLO READS to YOU

recognize sounds

Make shapes

Make letters

Sound out words

Explore more practice skills >

GROW A SKILL TOGETHER - PERFORM POETRY

I love writing poetry

Helpful Skills for Poetry:

COMPOSE A POEM

READ MASLO’s POEM

Play with sounds

Draw ideas

Articulate forms

Analyze old poems

Explore more practice skills >

Pragmatic Implementation - Maslo Embodiment

Expressive Boundary (UI/UX)

• Internal state
• Long Term Data Storage
• ML/Memory Models
• Signal Emission

• Signal Sensors
• Phone sensor access
• IoT add ons

• Manipulates
• Digits
• Extensions
• Limbs
• Wheels/Movers
• Embeddings

• Programmable materials (cloth, metals, leds, filament)
• Vibration patterns for rigid surfaces
• Phone and PC persona UI

Internal Memory

Signal Detection

Physical Body

Skills Template - Platform Extends Easily

Skills Template - Most skills will be default settings with tweaks. Skills are simply parameter settings on those four layers of Maslo platform.

Everything we do can be defined by a skills template. EVERYTHING.

Dying is a skill (crudely…)�Expression: Gray, desolve, disapparate and “delete” from hardware embodiment�Internal State: Persistence goes to 0, Training Type: Explicit�Sensor: Disconnect�Body: works in all forms

Arithmetic is a skill�Expression: calculator UI�Internal State: Persistence set to infinity for all numbers less than X, Training Model Type: Implicit (can learn from axioms/patterns)�Sensor: Vision, Sound, any really�Body: works in all forms

Big Points about Re-Framing all of Computation

Co-Learning with Maslo will replace COMPUTER PROGRAMMING.

Co-Learning with Maslo provides a sense of SHARED CONSEQUENCES

Co-Learning with Maslo makes AI learning tied to Human Concerns

Co-Learning with Maslo takes the best, wisest approaches of humanity and integrates computers

Co-learning with Maslo doesn’t add yet another consumptive activity to life.

Co-learning with Maslo reduces isolation and increases pro-social behavior.

Maslo.AI

Computational Being Platform�Aka general AI.

Experience live things:

https://un1crom.github.io/maslo/research/