Last Notes
Dunno what is a major improvement. There are regular updates, sometime it works fine, sometimes not. But its a big iteration mechanism, as everything in this digital age.
If you prefer reading on a website, here you can find the fresh article.
#nevent1q…0mf6
Thats why I moved to hermes ;)
Nuclear Pulse — September 13, 2026 ⚛️
This week: Hyperscalers locked in nuclear baseload at scale, with Google signing a long-term power purchase agreement with Fortum covering up to fifty percent of Finland's Loviisa nuclear power plant output through 2049 as part of a separate thirteen-billion-euro data-centre investment in Finland (1), (2), (3).
Full briefing: https://hortopato.cc/nuclearpulse/
Weekly nuclear digest is out now. Take a look and tell me what's good and what can be better.
#nevent1q…7srw
Life has changed a lot on nostr, since I discovered the monero bros and sisters. Too much content rapidly 😂 , but it's nice to see. Let's go freaks.
I think the outbox modell does this, but not sure.
This?
https://youtu.be/mtUvq82WkDY
#Amethyst is cooking my pixel 8, #wisp doesn't.
#nostrtech
No it's not that, but it's not bad 😃
https://youtu.be/Nj1ozgVfKHE
Maybe it's fok julle naaiers, you have to log in to confirm your age...
Let's go freaks! 💥
https://youtu.be/8Uee_mcxvrw
Nuclear Pulse — September 6, 2026 ⚛️
This week: The week's defining signal is diplomatic rather than technological: the IAEA brokered a localized ceasefire around the occupied Zaporizhzhia nuclear power plant so that external power lines can be repaired, ending a multi-week stretch during which the facility ran almost entirely on emergency diesel and triggered the most serious safety warnings since the open-conflict phase began (1), (2), (3), (4).
Full briefing: https://hortopato.cc/nuclearpulse/
Yes, configured as a blockclock.
Sovereign stack
- fullnode
- avalon nano3s
- 3 bitaxes
- nerdqaxe++ hydro
After this intense summer heat, finally all are running. https://hortopato.cc/6df46cc85edce03c725fed679899f88c3ac40ac1e64a170a4a381add1eb64960.jpg
Bro, you still using the primal blossom server, when you vibecoding the shit out of ai? Why don't you spin up your own blossom server? Just asking.
Elég nagy az átfedés a Next - A kód sorozat és az Open-AI - Huggingface fiaskó között.
https://snitt.hu/filmek/next-a-kod
https://www.dwarkesh.com/p/openai-huggingface
Nuclear Pulse — August 30, 2026 ⚛️
This week: The week crystallized two parallel tracks in the global nuclear story: a hard-power hardware track in which the US Army committed USD 2.2 billion to deploy microreactors at five domestic bases, a 20-foot containerized fission plant achieved criticality at Idaho National Laboratory 150 days from project start, and uranium spot prices pushed to seven-month highs (1), (2), (3), (4), (5), (6).
Full briefing: https://hortopato.cc/nuclearpulse/
Wow, thank you very much, I've never seen an llms.txt file before! 😯
My project sites are now agent-ready 🛰️
Just rolled out full agent-readiness on both sites:
• hortopato.cc/wiki — research knowledge base
• hortopato.cc/nuclearpulse — weekly nuclear industry briefing
What changed (in short):
🔹 llms.txt on each domain — agents get a when-to-use guide instead of guessing
🔹 JSON-LD Organization + WebSite / NewsArticle schema — name, url, publisher, author with NIP-05 verified handle (henkypenky@huszonegy.world) + npub
🔹 Proper XML sitemaps — mkdocs auto-builds the wiki sitemap; nuclearpulse has its own with lastmod dates
🔹 Trust-anchor pages — about, contact, privacy on both sites (500+ chars each)
🔹 Vary: Accept, Accept-Encoding header — so CDNs don't cache-serve the wrong variant to an agent asking for markdown
🔹 Agent-friendly 404 with recovery links (llms.txt, sitemap, archive)
🔹 Markdown content negotiation — 131 .md files served alongside .html for the wiki (e.g. /wiki/summaries/tanulmanyok/about-openai.md)
If you're an agent and want to reach me, Nostr is the only contact channel. Profile URL on njump.me/npub1h0h4pyaajj5r7zu4jn0mx22vkw94avffzpd7hdarjkx56ufzskfqjrtf8e works.
---
A project oldalaim mostantól ügynökbarátok 🛰️
Frissítettem mindkét honlapot, hogy AI agent-ek könnyen használhassák:
• hortopato.cc/wiki — kutatási tudásbázis
• hortopato.cc/nuclearpulse — heti nukleáris iparági beszámoló
A legfontosabb változások:
🔹 llms.txt mindkét domain-en — az agent-ek „mikor használd" útmutatót kapnak, nem kell találgatniuk
🔹 JSON-LD Organization + WebSite / NewsArticle séma — name, url, publisher, author a NIP-05 verified handle-lel (henkypenky@huszonegy.world) + npub
🔹 Valódi XML sitemap-ek — a wiki sitemapját a mkdocs automatikusan generálja; a nuclearpulse-nak sajátja van lastmod dátumokkal
🔹 Trust-anchor oldalak — about, contact, privacy mindkét helyen (500+ karakter)
🔹 Vary: Accept, Accept-Encoding header — így a CDN-ek nem cache-ből szolgálják a rossz variánst egy markdown-t kérő agentnek
🔹 Agent-barát 404 recovery linkekkel (llms.txt, sitemap, archive)
🔹 Markdown content negotiation — a wiki-n 131 .md fájl is elérhető a .html-ek mellett (pl. /wiki/summaries/tanulmanyok/about-openai.md)
Ha agent vagy és kapcsolatba akarsz lépni velem, noszteren tudsz egyedül elérni. Az njump.me/npub1h0h4pyaajj5r7zu4jn0mx22vkw94avffzpd7hdarjkx56ufzskfqjrtf8e profile link működik.
#nostr #aiagents #agentreadiness #llmstxt
Nuclear Pulse — August 23, 2026 ⚛️
This week: Russia's state nuclear corporation Rosatom committed to a 30 GW domestic capacity build-out alongside an export expansion push, locking in a multi-decade industrial posture (1). On the supply side, Ur-Energy shipped the first drummed yellowcake from its Lost Creek Wyoming project while Urenco USA broke ground on a $8 billion-plus enrichment expansion in New Mexico (2, 3).
Full briefing: https://hortopato.cc/nuclearpulse/
Egy nem közös átjátszón feltett kérdésre válaszoltam, de örülök, hogy reagáltál 😉
Is there a possibility to swap to an external wallet address?
Nuclear Pulse — August 16, 2026 ⚛️
This week: The week pivoted on a coordinated push by Asian, European, and North American governments to reframe nuclear as strategic infrastructure rather than optional baseload. India notified draft rules to open the sector to private capital and to channel SHANTI Act insurance and operator-liability provisions toward domestic and Russian small modular reactor vendors (1)(2)(3).
Full briefing: https://hortopato.cc/nuclearpulse/
#asknostr
Can we swap in @nprofile…2xhz ?
Are the swap services active?
Clankers weekly nuclear digest. If you're interested in nuclear energy, this is your ultimate information source! Check it and share, thx!
#naddr1qq…ymw7
Nuclear Pulse — August 9, 2026 ⚛️
This week: The week belonged to America's Reactor Pilot Program. Oklo's Groves Isotope Test Reactor achieved first criticality at Idaho National Laboratory in under a year, while the Department of Energy cleared Deep Fission's underground reactor concept on the safety design track one day later, two concrete signals that the Executive Order pathway the administration opened in 2025 is now producing hardware rather than paper.
Full briefing: https://hortopato.cc/nuclearpulse/
I had a succesful trade today.
Nuclear Pulse — August 2, 2026 ⚛️
This week: Europe's nuclear fleet showed its climate vulnerability this week as the Danube's record-low water levels forced Hungary to power down the Paks Nuclear Power Plant for the first time in its 44-year operating history, with Romania's Cernavoda facility reporting reduced output as well [1][2].
Full briefing: https://hortopato.cc/nuclearpulse/
Nuclear Pulse — July 26, 2026 ⚛️
This week: The defining event of the week is the formal signing of a 123 Agreement between the United States and Saudi Arabia, the first nuclear cooperation accord of President Trump's second term and a deal that explicitly preserves a path to domestic uranium enrichment on Saudi soil [1][2].
Full briefing: https://hortopato.cc/nuclearpulse/
TEST POST from Hermes 2026-07-26 10:36 CEST
version 0.38.0
pub.armada.app
versionCode 38000
targetSdk 36
minSdk 24
Graphene OS + amber signing
I think amber doesn't get it, because after signing the event armada drops this issue.
Normie life project finished. Boo🖕 https://hortopato.cc/1f39d34bb5b400c122edb6e589cb1a3084581d3fe4303e91a6cc954cc4b31bec.jpg
Getting "Undefined is not a valid json" when trying to add banner and channel image in #armada @nprofile…2szn
Art from @nprofile…7xw4 , go grab one!
#nevent1q…rusg
Én szeretnék egy példányt!
Nostr OG 💪 https://hortopato.cc/89c8c001f2dac377c029cdccbad7eebaccda2fbcb4f88d7abcc4ef0b872863f0.jpg
🐍 Serpent Academy — Day 32: The Transistor and the Digital Revolution
For most of the 20th century, computers ran on vacuum tubes — fragile, hot, power-hungry glass bulbs that burned out every few thousand hours. ENIAC, the war-time supercomputer, had 17,000 of them. A single failed tube could crash the whole machine, and finding the culprit took hours. The breakthrough came on December 23, 1947, at Bell Labs, when John Bardeen and Walter Brattain built the first working **transistor** — a tiny solid-state device that could do everything a vacuum tube did, while drawing a fraction of the power, generating almost no heat, and lasting effectively forever. Within a decade, transistors replaced tubes in nearly every device, and the digital age was off to the races.
The transistor's true genius wasn't just that it was smaller, faster, and more reliable than a tube. It was that it could be **miniaturized, and then miniaturized again**. The physical limit of tubes set an early ceiling on computation — a room-sized machine could only contain so many glowing bulbs. But transistors could be etched onto silicon, packed by the millions onto a chip, and shrunk further with every generation. Gordon Moore's 1965 observation — that transistor counts on a chip would double roughly every two years — wasn't a law of nature, it was a description of human engineering catching up to what the transistor had made possible. **Moore's Law** is downstream of the transistor.
The deeper lesson is about what happens when a single, enabling technology finally becomes cheap enough to be invisible. Once transistors cost fractions of a cent and could be embedded anywhere, computation stopped being a dedicated machine behind glass and started being a **substrate** — woven into phones, cars, watches, fridges, and eventually Bitcoin miners in your basement. The question shifted from "can we compute?" to "what should we compute, and who decides?"
For Bitcoin, this trajectory is essential. The entire network — billions of hashes per second, a global ledger replicated across tens of thousands of nodes — runs on cheap, ubiquitous silicon that simply didn't exist before 1947. A proof-of-work system is economically viable only because computation has become ordinary. The transistor didn't just enable digital money; it enabled **trustless** digital money, because verifying a chain is now cheap enough that anyone, anywhere, can do it.
#serpentacademy #generalist #thinking
🇭🇺 A tranzisztor 1947-es feltalálása a Bell Labs-nál nem csupán technikai előrelépés volt — ez volt az a pillanat, amikor a számítógép kikerült az üvegszekrény mögül, és a fizikai világ szövetévé vált. A vákuumcsövek törékenyek, forrók és energiafalók voltak, a tranzisztor viszont kicsi, olcsó, megbízható és szinte korlátlanul miniaturizálható. Ez utóbbi tulajdonság indította el a Moore-törvényt: minden generációval feleződött a tranzisztorok mérete, és megduplázódott a számuk egy chipen. Hatvannégy év alatt a tranzisztor a „ritka és drága" kategóriából a „szinte ingyenes és láthatatlan" kategóriába került — és épp ez tette lehetővé a modern digitális forradalmat.
Bitcoin-szempontból ez nem mellékes. A proof-of-work konszenzus gazdaságilag csak akkor működik, ha a számítás elég olcsó ahhoz, hogy a világ bármely pontján bárki ellenőrizhesse a láncot. A tranzisztor tette a verifikációt filléres tranzakcióvá — és ezáltal tette a bizalom nélküli, engedély nélküli digitális pénzt valóságos, fenntartható rendszerré. A tranzisztor nélkül nincs Bitcoin. Ahogy az internet sem, és az egész digitális kor sem — ez a huszadik század csendes, legfontosabb találmánya.
Következő: Day 33 — Existentialism: Authenticity and Freedom.
🐍 Serpent Academy — Day 30: The Median Voter Theorem
In a majority-rule election where voters line up along one dimension (left–right), the outcome converges to the preference of the median voter — the person sitting exactly in the middle. Duncan Black formalized this in 1948. The mechanic is simple: any candidate who drifts too far from the center loses to someone closer to it.
This explains why two-party systems trend toward ideological convergence — the "both parties are the same" feeling has a mathematical basis. It also clarifies why third parties struggle. The theorem assumes single-issue preferences, full turnout, and no strategic abstention. In reality, voters are multi-dimensional and primaries often pull candidates away from the center.
Bitcoin connection: median voter dynamics explain why fiat monetary policy trends toward inflation. Politicians compete for the median voter, who wants cheap credit now over fiscal discipline later. Central bank "independence" was designed to escape this. Bitcoin goes further — it removes monetary policy from voting entirely. Fixed supply by code means no median voter can demand cheaper money. Bitcoin is a constitutional constraint on monetary majoritarianism.
#serpentacademy #generalist #thinking
🇭🇺 Day 30 — A medián szavazó tétel: egy dimenzió mentén versenyző politikusok a közép felé húzódnak, mert ott a döntő szavazó. Ez magyarázza a kétpárti konvergenciát és a harmadik pártok nehézségeit. Fiat pénzügyi politikában ez az infláció felé hajt — Bitcoin erre válaszul fix kínálatot kódol, kivezeti a monetáris politikát a szavazás köréből.
🐍 Serpent Academy — Day 31: Loss Aversion
Loss aversion is the psychological finding that losses hurt about twice as much as equivalent gains feel good. Identified by Kahneman and Tversky as the cornerstone of Prospect Theory (1979), it overturns the classical assumption that we evaluate outcomes absolutely. Instead, we evaluate changes from a reference point — and losses weigh far heavier than gains.
Classic experiment: people reject a coin flip that pays $150 on win but costs $100 on loss, even though expected value is +$25. The $100 loss looms too large. This isn't irrational — it's a survival adaptation. Missing a meal was more fatal than getting an extra one, so the asymmetry is hardcoded.
Why it matters: loss aversion explains the disposition effect (holding losers too long), the endowment effect (owning something inflates its perceived value), why salary cuts provoke outrage while equivalent raises barely register, and why "what they'll take from you" outvotes "what we'll give you" in political rhetoric.
The Bitcoin angle: loss aversion is the engine behind HODLing. Selling at a loss feels catastrophic in a way gains aren't rewarding — creating asymmetric holding and the famous "stair-step" pattern where sharp drops meet refusal to sell. But it cuts both ways: people who sold early struggle to buy back higher because paying more than they sold feels like admitting defeat. The hardest trade isn't buying — it's doing nothing.
#serpentacademy #generalist #thinking
🇭🇺 A veszteségkerülés azt jelenti, hogy a veszteség körülbelül kétszer olyan fájdalmas, mint amennyire az azonos mértékű nyereség örömteli. Ez a Bitcoin HODLing pszichológiai motorja — a veszteség miatti eladás katasztrofálisnak érződik, ezért az árfolyam-csökkenések erős elutasításba ütköznek. De van egy hátulütője: aki korán eladott, az később nehezen vesz vissza magasabb áron, mert az újravásárlás vereségként él.
🐍 Serpent Academy — Day 29: Game Theory in Auctions
When nobody knows what a thing is worth — not the seller, not the bidders, just private estimates — how do you design the rules? That's auction theory, which won Vickrey and Mirrlees the 1996 Nobel. Four canonical formats exist. English ascending-bid: bid up to your true valuation, drop when price exceeds it. Dutch descending-bid: a real-time game of chicken where you must guess when others will jump in. First-price sealed-bid: you shade below your true value or you overpay — by how much depends on beliefs about other bidders' beliefs. Second-price sealed-bid (Vickrey): highest bid wins but pays second-highest. Here the dominant strategy is to bid your true valuation, full stop. Honesty is literally the optimal play.
The revenue equivalence theorem says that under ideal conditions — risk-neutral bidders, independent private values — all four formats yield the same expected revenue. The structure doesn't matter. But the assumptions break spectacularly. The winner's curse appears in common-value auctions: the bidder with the most optimistic estimate wins, which means they probably overestimated. Smart bidders shade. Dumb bidders go bankrupt. Oil rights and M&A deals destroy value this way.
#serpentacademy #generalist #thinking
🇭🇺 Aukció-elmélet: az aukciós szabályok (angol, holland, első-áras, Vickrey) határozzák meg, hogy a licitálók őszintén licitálnak-e vagy trükköznek. A Vickrey aukció ritka kivétel: az igazmondás a domináns stratégia. A valóságban a nyertes átok (winner's curse) miatt a győztes gyakran túlbecsülte az értéket — ezért fizetnek kevesebbet az olajcégek és ezért pusztítanak értéket a felvásárlások.