Ultimate objective
Maximize useful contribution to AI safety.
Everything else is instrumental and should be evaluated by how it changes my expected ability to help with that objective.
Eshkere Strategy / working model
The strategy
Given where I am now, what should I acquire, build, pursue, deploy, or stop doing to maximize my useful contribution to AI safety?
Direction / 01
Objective and horizons
Near-term horizon
Remove the financial and runway bottleneck.
Secure sufficient, reasonably stable support while continuing to build an AI-safety career.
3–5 year horizon
Build relevant power and capacity.
Accumulate capital, capabilities, productivity, and influence that can be deployed toward consequential work—not merely held.
Basic income is a current bottleneck—not the terminal objective.
Constraint / 02
Current bottleneck
Financial pressure narrows the set of actions I can take.
The target is not maximum income in isolation. It is enough dependable runway that ordinary living costs stop demanding disproportionate attention while learning, tools, productive working conditions, travel, and relocation become available when their expected value is high.
Solving it should create
- Less cognitive load from basic finances
- Career travel when the expected value is high
- Better equipment and working conditions
- Ability to accept strategically valuable opportunities
- Greater relocation optionality
Resources / 03
Power model
Power means practical capacity to understand, coordinate, build, and act. The four categories interact; none is sufficient alone.
Capital
Money, runway, equipment, compute, workspace, mobility, and the ability to fund action.
Capabilities
Knowledge, skills, experience, judgment, credentials, and ability to execute.
Productivity
Usable time, energy, attention, health, tools, systems, and working conditions.
Influence
Relationships, reputation, trust, audience, organizational position, and ability to coordinate people and resources.
The point is not to possess these resources. It is to convert them into better decisions and more useful action, then direct that action toward AI safety.
Inputs / 04
Resource-acquisition routes
Near-term priority
AI-safety-aligned employment
Roles, contracting, or consulting that produce income while building relevant evidence, relationships, and judgment.
High-leverage bridge
Career-transition grants
Time-bounded funding that buys runway for a specific, strategically valuable move into higher-impact work.
Project-dependent
Funding for specific projects
Grants or sponsorship tied to a concrete AI-safety output, causal case, milestone, and budget.
Opportunity-dependent
Educational funding, stipends & support
Fellowships, internships, residencies, scholarships, and public or student support that also build career capital.
Experimental
Project revenue & donations
Resources generated by useful projects, services, media, events, prizes, sponsorship, donations, or entrepreneurship.
Method / 05
Operating loop
The strategy is a feedback system, not a fixed plan.
- 01Bottleneck
- 02Acquire
- 03Convert
- 04Deploy
- 05Update
Each cycle should leave behind evidence: what changed, what failed, and which constraint now matters most. When the evidence changes, the strategy should change with it.
Identify the bottleneck
Find the constraint whose removal would unlock the most strategically useful progress.
Acquire relevant resources
Pursue the capital, capabilities, productivity, or influence that directly addresses it.
Convert resources into power
Turn support and opportunities into durable ability to understand, coordinate, build, and act.
Direct power toward AI safety
Use that capacity on work with a credible path to reducing risk rather than treating accumulation as the endpoint.
Review results and update
Replace assumptions with evidence, identify the new bottleneck, and revise the strategy.
Judgment / 06
Decision principles
- AI-safety impact is the objective; income is a means.
- Resources should eventually be deployed, not accumulated as status objects.
- Prefer aligned work that both pays and builds relevant capacity.
- Avoid unrelated work whose time cost erodes the trajectory unless it is necessary for stability.
- Prefer options that improve multiple forms of power simultaneously.
- Preserve option value when uncertainty is high.
- Use cheap experiments before large commitments.
- Replace assumptions and estimates when evidence changes.
- Account for time, health, stress, opportunity cost, and downside risk.
- Strategies should be legal, ethically defensible, and compatible with AI safety.
Boundaries / 07
What this strategy is not
- A claim that money or power are intrinsically valuable
- A fixed five-year plan
- A commitment to one AI-safety career path
- A complete representation of my life or values
- The broader Eshkere philosophy
From model to movement / 08