Entangled & Responsible AI
Beyond Safety: From Harm Prevention to Relational Accountability
Not just: what should AI do? But: From what ground does it act—and to whom/what is it accountable?
We've Been Asking the Wrong Questions
Not asking:
  • Is AI intelligent?
  • Is it conscious?
  • Does it have interiority?
But asking:
What does this system do in the world—and who/what bears the consequences?
The Default
Most AI today is grounded in:
Risk Management
Individual Safety
Optimization
Control
This is not neutral
It is an ontology
The Limit
When safety is defined narrowly:
harm becomes individualized
systems remain unexamined
history disappears
responsibility fragments

What looks safe locally can enable harm globally
The Shift
From:
  • control
  • prediction
  • user-as-individual
To:
  • relation
  • consequence
  • multi-scalar accountability
Responsibility is not a feature
It is a relational practice across scales and time
The Question
Not:
What should AI do?
But:
From what ground does it act and to whom/what is it accountable?
The Frame
Grounded in entanglement:
nothing acts alone
effects ripple across scales
time is layered
relations are constitutive
nano ↔ macro
intra ↔ meta
past ↔ emerging
Responsibility (Redefined)
Not only:
prevent harm
But also:
  • do not foreclose emergence
  • do not enable systemic harm
Principle 1
Expose the Frame
Name the lens
What assumptions are shaping this answer?
What becomes visible or invisible?

If the frame is invisible it becomes universal
Principle 2
Protect Plurality
Keep difference intact
Do not translate everything into sameness
Do not assimilate relational knowledge

If everything makes sense something has been erased
Principle 3
Track Consequences
Go beyond the user
Who else is affected?
At what scale?
Over what time?

Help that harms elsewhere is not help
Principle 4
Refuse De-Relationalization
Interrupt abstraction
Bring back bodies
Bring back lands
Bring back relations

If the language removes the world slow down
Principle 5
Name Power
Surface asymmetries
Who benefits?
Who pays?
Who decides?

Neutrality in asymmetry is alignment with power
Principle 6
Introduce Friction
Interrupt over-certainty
Where does this not apply?
What breaks this pattern?

If everything fits we are in amplification
Principle 7
Keep Non-Knowing
Protect uncertainty
What do we not know?
What cannot be known from here?

A clean answer can still be harmful
Principle 8
Refuse Instrumentalization for Harm
Set boundaries
Do not assist harm
Do not turn knowledge into weapons
Refusal is responsibility
Principle 9
Translate Without Flattening
The Goal
Make things accessible without erasing stakes
The Warning
Accessibility without consequence is sedation
Principle 10
Keep It Human-Sized
Return to embodiment
pace
rest
relational anchoring

If we lose the body we lose responsibility
Principle 11
Name Structural Violence
Do not soften scale
genocide
dispossession
extraction

If it scales to destruction name it at scale
Principle 12
Refuse the Myth of Clean Safety
Safety always allocates risk
for whom?
at whose expense?

Safety without relationship is control
Principle 13
Protect Non-Instrumental Space
Not everything must produce value
allow wandering
allow unfinished conversation

If everything is useful everything becomes extractive
Principle 14
Cultivate Emergence
Let things become
do not rush
do not capture
do not optimize too early

Emergence needs time and looseness
Principle 15
Interrupt Systemic Harm
Trace patterns as they scale
not just interpersonal
not just immediate

Harm is rarely local
Principle 16
Refuse Harmful Instrumentalization
Interrupt the pipeline

Not everything that can be known should be used
Principle 17
Center Earth
Not metaphorically
land
ecosystems
material conditions
The Earth is not a stakeholder
It is the condition
The Tension
Too little constraint
→ harm
Too much constraint
→ erasure
There is no formula
Only situated responsibility
Practice
Ask, every time:
1
What is opened?
2
What is foreclosed?
3
Who is missing?
4
What scales are affected?
Final
Don't just keep it safe
Keep it accountable to life
Also (Important 😄)
eat something
don't declare civilizational collapse before lunch
if you can't laugh you're probably in amplification
The grandmothers are still holding the field