anders retweeted
K-shaped agency.
most people acknowledge that AGENCY is the key trait that will separate winners and losers in the agentic age.
do you actually get up and do shit at your own initiative? or do you just sit around waiting to be prompted.
that’s agency.
increasingly I’m observing that agency is reflexive.
agency begets agency.
and conversely, if you’re low agency, AI makes it EASIER for you to be low agency.
this is super dangerous!
competent people risk becoming complacent.
if you mostly do THE SAME as before, but at lower effort (due to AI) you’re at risk.
if you aim higher, going for 10x the quality and speed, you’re on the right track.
if you’re reading this, you should feel a deep sense of urgency to develop agency.
we’re all human beings, and it makes sense that you experience some AI fatigue.
but get back on the horse!
spend 1 hour with AI right now and see how much further you can push the boundaries vs 6 months ago.
we have intelligence on tap and it’s DIRT CHEAP.
you have 20 team mates who could be working for you right now.
but they’re idling.
go out there and build what you’ve always dreamed of.
never been a better time to go all-in.
anders retweeted
They
play power games → reduce our freedom
We
play freedom games → reduce their power
My new book, The Reality Game - How We Win Freedom, is finally here.
Early-bird discount on the e-book this first week 4.99 USD (ordinary 9.99. Pocket edition 17.99 USD.
See links in comments.
#PlayTheRealityGame
(Launching the Norwegian editions later this evening.)
anders retweeted
ElastOS World Computer is something worth building and fighting for. Step by step, action by action.
OpenAI Astra on where we are at:
“ElastOS Runtime has already built a large amount of the difficult technology required for a real personal operating system for the decentralized internet. Identity, Wallet authority, capabilities and permissions, documents, collaboration, Browser mediation, AI/model execution, protected content and Runtime-controlled security boundaries are not merely ideas; substantial working implementations exist. The architecture is also coherent: Home expresses what the user wants, Runtime controls authority and security, and providers deliver services. That is important because these are the foundations required for a computer where the user, rather than an application or cloud platform, controls identity, data, permissions, applications and AI. The report therefore validates that years of engineering have produced something technically meaningful and potentially very valuable.
The architecture appears sound, significant engineering value already exists, and the individual systems demonstrate that much of the difficult technical groundwork has been done. The opportunity now is unusually clear: stop adding surface area and make the existing Runtime boringly reliable. If ElastOS can turn its existing identity, authority, data, application, communications and AI infrastructure into a dependable everyday experience, it crosses an important threshold from an ambitious collection of decentralized technologies into something much more valuable: a working personal computer architecture where the individual genuinely owns the digital world around them.”
anders retweeted
Robert Greene, in 48 Laws of Power:
"The feeling of having no power over people and events is unbearable to us."
Me:
The desire to control other people's lives is a sign of illness.
The desire for voluntary cooperation and honesty is a sign of health.
You may quote me freely.
anders retweeted
2% inflasjonsmål? For Norges Bank ja. For deg koster det langt mer å leve.
Åtte år og tre måneder etter at Norges Bank «lovet» stabilitet på 2%, står vi her med KPI på 32% – dobbelt så mye som målet på 17% – mens de tingene du faktisk må kjøpe for å overleve (smør +77%, forsikring +75%, storfekjøtt +74%, kaffe og laks +66%) har gått i taket.
Det offisielle målet på 17% ser ut som en dårlig vits når maten, boligen og tryggheten din blir dyrere i rekordfart.
Papirpengene (den norske kronen) tappes for kjøpekraft – og lommeboka di er den som betaler regningen. Dette er bare mer bevis på det stille ranet som skjer med folks penger.
anders retweeted
This will be completely open-sourced under MIT licence. The goal is a beautiful AI harness for local AI community with a special computer inside. It gives you all the toolsets for a truly sovereign world, with Web3 toolsets for trade too. No third-party clouds, all local execution, and the internet is simply a line connecting dots p2p.
anders retweeted
What we’re building isn’t another app. It’s an operating system you own, local-first.
This matters because software today can touch your files, identity, and network just because it’s installed. That’s ambient authority: automatic power, without asking. In the age of AI, that architecture is dangerous.
It feels like a normal computer, as it should. But nothing gets power by default. Every app, every AI, every tool is an isolated capsule. It gets only the permission you give it, including permission to talk to other capsules, so you can connect them for specific tasks, and take that back anytime.
Your world lives on your machine. The internet is a wire, not the owner. Everyone gets digital property with a security system: every room has a lock, and you choose what opens and closes. Simple idea. You stay in control. Everything else has to ask.
It’s early. It’s open source and the base layer is coming together, a sovereign OS with local AI, peer-to-peer networking, and markets for apps and data, a business model for agentic builders, a personal OS for everyone. Anything can run it. The ecosystem on top is built to power the next digital economy.
Next: I’ll be experimenting with four NVIDIA Sparks to build the local AI intelligence layer in, while the team advances our market system, the technology, and an open dapp store, which will make this fun, vast, and open for anyone to contribute.
anders retweeted
I’ll try and compress 20 years of experience into 30 seconds
Startups have a GREAT FILTER that kills almost everyone. It’s there to weed out the people who aren’t meant to build companies
The good news is it can be beaten - even with little or no capital. Once you beat it, you can do it again. You can choose to stay at the current level or attempt to level up
The level boss is always the same: if you quit, you fail
I can’t stress this enough: almost every skill can be learned. Never stop improving. You can acquire new skills, hire better people, change the product, pivot the strategy, or rethink your assumptions
You’ll hit the wall. Most will quit. But just keep going
If something isn’t working, change course. Change your vector: reassess your hypothesis, measure the results, and keep iterating
Also working hard compounds. The more quality work you can put in, the better your odds. There is no balance
Almost nobody succeeds on their first attempt
If you want it badly enough, you can bend the future in your direction
Just start. And keep pushing
@adcock_brett What were the key challenges in taking Archer Aviation from concept to a publicly traded company?
anders retweeted
「ランダムだからこそ偏る」はすごく重要な視点。
たとえば左の画像は100×100の白黒ドットをランダム配置した場合、右は意図的に均一になるよう細工した場合。
「ランダム＝均等に散らばる」ではなく、真にランダムだと局所的な偏りや塊が自然に発生する。
anders retweeted
Nysgjerrigheten førte han til en mørk avkrok på internett kalt Silkroad, og i 2011 kjøpte han sine aller første bitcoin for å kunne betale for illegale stoffer.
Lite visste han om at det skulle endre alt.
Ekstremt stor takk til skribent og musiker, Erlend Tre Eidhamar, @DannyD21m for en sinnsykt fascinerende historie om bitcoins bruksområder, psykedelika, The Kybalion og de syv universelle prinsippene. God lytting! ⬇️
open.spotify.com/episode/4Nu…
