my personal operating principles for 2025
version 1; updated: 2025-02-02
It’s a mistake to let an uncertain future ruin the present
— Ilya Sutskever (@ilyasut) August 13, 2023
This document was written for myself.
For a while I was struggling with anxiety while thinking about the myriad decisions and choices coming my way in the next few years. The main contributors to this anxiety were: 1. a general cloud of uncertainty about the AI boom (30%), 2. uncertainty aboÏut the future of software engineering as a career (30%) 3. current geopolitical climate (5%) 4. fear of DOOM from AI (10%) 5. 25% others.
Here, I synthesize a set of principles to guide me through the upcoming years. They should be revisited regularly (see [[#5. recalibrate relentlessly]]).
1. build an antifragile core
“build resilience first; brilliance follows. – o3-mini”
safeguard your machine: Treat your body and mind as non-renewable capital. Prioritize sleep, exercise, and stress management like you would any critical infrastructure.
adopt a barbell strategy: Anchor your life in low-risk fundamentals—health, financial reserves, relationships—while taking calculated, asymmetric bets on
<redacted>
.practice good cogsec AI slop is everywhere, algorithms manipulate and influence your thoughts. The mind is a pickle jar – it has limited capacity and you must be selective about what goes in.
avoid lifestyle creep: Keep your annual burn in check and leverage favorable financial instruments (like a
<redacted>
) to preserve liquidity.maintain life optionality: Consider strategies (citizenship, real-estate, relationships) to secure long-term mobility in a dynamic global landscape.
why?
AI-driven volatility is inevitable. By building an antifragile core, you prepare to ride the waves rather than be overwhelmed by them.
2. act with 10x agency
“You can just do things – X”
if you value intelligence above all other human qualities, you’re gonna have a bad time
— Ilya Sutskever (@ilyasut) October 7, 2023
play at a higher level When the rules change, stop playing the old game.
- If AI is automating SWE tasks, don’t fight the current—swim upstream.
- Architect > Coder. Systems thinking, AI oversight, and security aren’t getting one-click replaced anytime soon.
- Learn how to prompt, steer, and verify AI-generated output—fast.
- 💡 Mull on this: SWEs are competing with AI + AI-literate SWEs. Be the latter.
don’t just consume, ACT: Invest time in micro-projects (under 100 hours) that explore the intersection of AI and X. /s now is the time for action /s
If you’re not using AI daily, you’re already behind. Work with AI to amplify your actions. Proactively reposition yourself into roles where AI and security converge.
why?
Uncertain futures get tamed by tangible output, not Twitter threads or Medium essays.
3. engage with SPACIOUS PASSION
“By passion the world is bound, by passion too it is released – Hevajra tantra”
TODO: This section is a work in progress.
be earnest:
Welcome each experience fully without clinging to fixed outcomes. Let curiosity drive you to explore diverse perspectives and opportunities.be a scout, not a soldier:
- REALIZE THAT TRUTH ISN’T IN CONFLICT WITH YOUR OTHER GOALS
- LEARN TOOLS THAT MAKE IT EASIER TO SEE CLEARLY
- APPRECIATE THE EMOTIONAL REWARDS OF SCOUT MINDSET
- balance intensity with spaciousness:
Combine deep commitment with a flexible, expansive mindset. This balance enables creative and effective responses to life’s challenges.
why?
engaging in this manner can result in mastery, power, playfulness, and nobility.
4. prioritize what compounds
“Be good to your future self by prioritizing what compounds: capital investment, relationships, audience building, discipline, and self-reflectiveness. - Julian Shapiro”
capital is optionality:
Building equity and keeping burn low creates lasting financial leverage.cultivate deep relationships:
Focus on high-trust, long-term relationships instead of superficial transactional networks.value family:
View family, children and close relationships not as costs but as principal-protected investments with multi-decade returns.develop genuine skill capital:
As AI commoditizes intelligence, focus on 1. fundamentals 2. ability to learn new fields of knowledge quickly (Hamming). Hone system-level thinking, architecture, and security oversight—abilities that remain rare and highly valuable.MOVE FAST It compounds faster than you realize.
why?
Durable assets—financial, relational, or skills—compound over time and provide a robust hedge against rapid change.
5. recalibrate relentlessly
“The map is not the territory. Update, don’t ossify.” “adaptability > anxiety”
nobody should give or receive any career advice right now. everyone is broadly underestimating the scope and scale of change and the high variance of the future. your L4 engineer buddy at meta telling you “bro cs degrees are cooked” doesn’t know shit
— roon (@tszzl) December 23, 2024
conduct regular reality checks:
Every three to six months, assess if AI is accelerating or stagnating and whether your skills and role are evolving accordingly.stay fluid and avoid fixed mindsets:
Recognize that AI’s progress is lumpy rather than linear; remain open to pivoting your strategies as new trends emerge.
why?
In a rapidly changing environment, the only real mistake is freezing in place—adaptation is your safeguard against obsolescence.
Credits
Inspired by theZvi, Julian Shapiro, Paul Graham, Ilya Sutskever, Bryan Johnson & Don’t Die, vividness.live, roon, Nassim Taleb, Nick Cammarata, Sam Altman, Julia Galef, O3, r1.
- https://benjamintodd.substack.com/p/how-can-an-ordinary-person-prepare
- https://www.lesswrong.com/posts/KFFaKu27FNugCHFmh/by-default-capital-will-matter-more-than-ever-after-agi
- https://stratechery.com/2025/ais-uneven-arrival/
- Maxwell Tabarrok:
The four futures for cognitive labor:
Like mechanized farming. Highly productive and remunerative, but a small part of the economy.
Like writing after the printing press. Each author 100 times more productive and 100 times more authors.
Like “computers” after computers. Current tasks are completely replaced, but tasks at a higher level of abstraction, like programming, become even more important.
Or, most pessimistically, like ice harvesting after refrigeration. An entire industry replaced by machines without compensating growth.