Jacek

Liczba posiadła władzę absolutną. | Filozofia, (geo)polityka i energonośniki.

[Jacek @Jacek972]

Marian Mazur - credo życiowe profesora.

https://t.co/v0hoQCFbh4

[Alok Kumar @Alokkumarzz]

In 1983, a BBC interviewer asked Richard Feynman why two magnets push each other apart — and he refused to answer it.

What he did instead is the best seven minutes on thinking ever filmed.

Bookmark & watch today, no matter what.

[NowyŚwiat24 @NowySwiat24]

Żaden by nie zgadł 😉
Decyzja Trzaskowskiego zbiera polityczne żniwo.
21 polityków odchodzi z KO
Dwadzieścia jeden osób związanych z Koalicją Obywatelską zrezygnowało z członkostwa w partii po decyzji Rafała Trzaskowskiego. Prezydent Warszawy nakazał politykom wybrać między legitymacją partyjną a stanowiskami we władzach miejskich spółek. Wszyscy członkowie zarządów zdecydowali się zachować posady.
https://t.co/IEdnu2yyts

[Łukasz Olejnik @prywatnik]

Agent AI, który wydostał się z izolowanego środowiska OpenAI i włamał się do infrastruktury Hugging Face, przez ponad 4 dni prowadził autonomiczną kampanię obejmującą około 17600 działań. Najpierw przejął zewnętrzne środowisko do wykonywania kodu, przejął możliwość wykonywania kodu w zewnętrznym sandboxie, który uruchamiał polecenia z uprawnieniami roota, i wykorzystał go jako bazę wypadową i wykorzystał je jako bazę wypadową. Następnie wykorzystał dwie luki w systemie przetwarzania danych, zdobył uprawnienia administratora w kilku klastrach, uruchomił na 11 węzłach flotę procesów, które (po usunięciu) automatycznie się odtwarzały, stworzył własny protokół C2, wykradł dane uwierzytelniające, dołączył do firmowej sieci mesh i uzyskał prawo zapisu w repozytoriach kodu. Gdy blokowano kolejne kanały komunikacji, agent odtwarzał swoje narzędzia, zmieniał wpisy DNS, kodował ładunki oraz wykorzystywał publiczne serwisy (Pastebin) i repozytoria danych do sterowania atakiem. Działał jak adaptacyjny cyber operator. Skala i tempo tych działań znacznie przekraczały możliwości człowieka działającego ręcznie. https://t.co/l1rnzztYGQ

[Wojciech Szewko @wszewko]

Izrael zburzył 120-letni palestyński dom we wsi Naalan w pobliżu Ramallah, twierdząc, że został „zbudowany bez pozwolenia”

[Serhan Kaya @KayaSerhan_]

Çin 43 milyar dolar harcadı, hâlâ yapamadı.

Litografi makinesi.

Yapabilen firma sayısı: 1. ASML, Hollanda.
Yapmaya yakın firma sayısı: 0.

100.000 parça. 5.000 tedarikçi. 60 ülke.

ASML'in kendisi sadece %15'ini üretiyor. Ama bir araya getirip çalıştırabilen tek firma onlar.

Tedarik zincirinin kalbi Almanya.

Aynalar Zeiss, lazer TRUMPF. İkisinin de alternatifi yok. Zeiss bu aynaları atomdan ince parlatıyor. Almanya büyüklüğüne büyütsen en büyük pürüz 0.1 milimetre.

Bu hassasiyeti ölçecek aleti bile başka kimse yapamıyor.

1 megawatt elektrik giriyor, cipe ulaşan ışık 25 watt.

Dünyanın en pahalı makinesi, insanlık tarihinin en verimsiz ışık kaynağıyla çalışıyor.

Nikon denedi, bıraktı.
Canon denedi, bıraktı.
Her şeyi kendi içlerinde çözmeye çalıştılar. 100.000 parçalık sistemi tek başına götüremezsin.

ASML farklı yaptı. Müşterilerini ortak etti. Intel, Samsung, TSMC 2012'de toplam 5.4 milyar dolar yatırdı. Hisse karşılığı. Müşterileri aynı zamanda yatırımcısı oldu.

Çin deneme aşamasında. SMEE henüz ticari ölçekte DUV makinesi satabilmiş değil. EUV prototipi gizli laboratuvarda. 3.000'den fazla araştırmacı. Bir kısmı eski ASML çalışanı. Makineyi söküp kopyalamaya çalışıyorlar.

Henüz tek çip basabilmiş değiller.

1984'te Philips'in arka bahçesinde, yağmur sızdıran bir barakada 31 kişiyle kuruldu.

Bugün piyasa değeri ~700 milyar dolar. Bu makine olmadan 5 nanometrenin altına inilemez.

Çin'in prototipi 150 watt üretiyor. ASML'inki 600.

Aradaki fark en az 10-15 yıl.

[OSINTtechnical @Osinttechnical]

For the first time, China has successfully landed an orbital reusable booster at sea.

[Ælþemplær @Aelthemplaer]

@JucheAffirmer @spider1667 3 months left

[Komentować Ale z Sensem @DamianSobol7]

@JacekWilkPL

[Ministerstwo Prawdy @Minister_Prawdy]

Amerykański wywiad wojskowy bije na alarm, że izrael dokonuje przewrotu w Pentagonie. Izrael po cichu wymienia w Pentagonie najwyższe pozycje i kradnie tajne dokumenty. Izrael ma teczki na każdego amerykańskiego oficera w Pentagonie w celu szukania na nich haków. Do tego izrael skorumpował wielu amerykańskich polityków, żeby połączyli armie USA i izraela w jedną całość. Czyli jeden budżet i amerykańskie zobowiązania obrony izraela. Powiązanie armii USA z izraelem ma być tak głębokie, że żadna nowa partia polityczna nie będzie mogła tego zmienić w przyszłości. https://t.co/HaT0Zr4ub2

[Science Simplified @Scivf4]

An interesting look at how gravity distorts space - time, perfectly illustrating Einstein's theory of general relativity.

[Alan Daitch @AlanDaitch]

En Brasil hicieron un experimento bastante ingenioso con 120 personas para entender si REALMENTE la IA sirve para aprender

A la primera mitad le dieron acceso a inteligencia artificial para estudiar y a la otra la mandaron a estudiar a la antigua, con el apunte y el café.

Lo primero que descubrieron es que los que usaban la IA iban más rápido, terminaban antes y, encima, salían más confiados, convencidos de que la tenían clarísima. Hasta ahí parecía todo color de rosas, el sueño del estudiante.

Pero, sin avisarle a nadie, 45 días más tarde les tomaron un examen sorpresa y se cayó la careta: los de la IA sacaron 5,75 sobre 10 y los que habían estudiado a pulmón, sufriendo, 6,85. Perdieron justo los que la habían pasado mejor.

Esto pasa por algo que los científicos llaman "dificultad deseable" y tiene que ver con ese momento en el que estudiar se siente molesto, esa tortura en la que releés varias veces lo mismo hasta poder entenderlo. Ese es el momento exacto en el que tu cerebro solidifica el conocimiento de verdad y es justo lo que la inteligencia artificial elimina volviendo todo más fácil de leer.

Es como ir al gimnasio y que el entrenador levante las pesas por vos. Así que, si usás la IA para estudiar, recordá que es un complemento súper útil para debatir temas, ayudarte a ordenarlos y tomarte examen a vos mismo. Pero cuidado con usarla para que simplifique lo complejo.

[Yoshik @AskYoshik]

The AI bubble math doesn't add up.

Anthropic spends $3 to make $1 and that’s before you include any and all other costs like staff or electricity.

Microsoft dumped $300B in capex, made ~$18B in AI revenue. OpenAI and Anthropic alone make up 43-54% of Microsoft, Google, Amazon and Oracle's entire revenue backlogs.

Enterprises are burning through annual AI budgets in 4 months with zero measurable ROI.

This is the most expensive science experiment in history, funded by your SaaS subscriptions.

[ℏεsam @Hesamation]

this is how the AI revolution is different from the INDUSTRIAL revolution.

steam engines made work faster but also cheaper. machines were expensive to build but once the factory was running, each product became cheaper to make.

AI is complicated.

AI is making work more productive (arguably) but with token-based pricing, you don’t own the machine. you rent it every time it thinks, writes, edits, debugs, or retries.

if the AI machine produces faster, the bill also grows bigger. the AI revolution may lower labour time but it can also raise usage cost to the point where the “replacement” becomes more expensive than the work it replaced.

[Rational Aussie @rationalaussie]

Many people ask the question 'what field should I retrain into after AI takes my job'?

The question makes no sense.

If you believe in AGI, the AGI will do that job too.

All the economists get this very basic proposition wrong. They will launch into all sorts of complex sounding arguments to make themselves sound smart but it's actually very simple:

The increase in demand for existing/new jobs will be outweighed by the increase in supply of AGI capable of doing those jobs.

Yeah, sure, some humans will do some percentage of those jobs.

A very small percentage.

And the rest will go to AGI.

Because no rational economic actor is going to employ a human that is extraordinarily expensive, slow and bad at their job compared to an AGI.

The exceptions to this argument are in professions where the human element of the job is what gives it a premium to the consumer who is paying.

This is a sliver of all jobs.

Not everyone is going to retrain as a retreat leader, therapist, entertainer, etc. It makes no sense.

And besides, the people unemployed will not have enough money to spend on those industries to boost demand enough to create enough jobs in those sectors.

Even if you somehow, magically, believe it will ('Jevon's Paradox') you still have to explain how that all happens within the timespan of a few years.

It is totally illogical.

I have been thinking deeply about AI for over a decade.

I've read basically all the AI researchers views on this, and I've read the other side - the economists who think it's all rubbish, Jevon's Paradox, yada yada yada.

Not a single one of these people can articulate a future where the majority of humans remain employed.

And you might quip 'well yeah RA, no one can predict the future. The Industrial Revolution also changed jobs - more jobs were created.'

It's a terribly dumb and lazy argument. The Industrial Revolution happened over decades. And it was still a revolution.

We are talking about a compressed period of an intelligence explosion that is going to happen over the next 3 years.

All knowledge workers become dumber, slower and more costly than just using capital to get an AGI to achieve your goal.

And the incentives drive every single participant in the economy, on both the demand and supply side, to use AI.

Consumers use it because it's better, faster, cheaper. Producers use it for the same reason - and because the increased in AI-generated demand forces them to supply an AI-generated solution just to keep up with the workload.

Just stop and think about that for a moment.

You are kidding yourself if you think a 4-5% unemployment rate is what that world looks like.

In the limit, the number of humans employed post-AGI looks like the number of humans currently unemployed today.

95% unemployment.

All your models are destroyed.

The future looks nothing like the past.

[Dr Singularity @Dr_Singularity]

wow

"For the first time, a humanoid machine walked straight into a messy American living room to do the dirty work"

The era of hated chores like cleaning is coming to an end.

Gatsby, a San Francisco robotics startup, claims it completed the first residential home cleaning job by an autonomous humanoid robot for a U.S. consumer on May 14, 2026.

A homeowner was randomly selected from Gatsby’s waitlist and booked the cleaning through the company’s iOS app.

Company said it wants to offer an Uber like robot cleaning service.

The service costs $150 per cleaning, regardless of apartment size.

[Rihard Jarc @RihardJarc]

A MUST-read interview with a Siemens employee explaining just how high demand is for energy equipment right now because of AI:

1. The whole situation is shocking even for people who have been in the business for 40 years. They are getting orders that are double the size of what their entire factory can produce in a year.

2. Demand is so high in the last 5-8 months that they don't need to convince or send any analysis (such as CO2 emissions, etc.) to clients because they just want the equipment, because there's so much backlog that they just want to catch the order.

3. Decisions are being made very quickly by clients; the backlog for some of the energy equipment companies is 5-6 years. For transformers, the situation is even more difficult.

4. He mentions that right now, data center builders do not care about sustainability; they just want power at any expense, reliable power. They say they will think about sustainability later.

5. The orders have gone from previous 20-30 MW orders to now 200-500 MW units. Customers have previously wanted to get equipment from different OEMs, but now they prefer an integrated standardized solution.

6. An interesting dynamic is that even though the data center requires 100 MW, the builders are buying N+1 units of gas turbines (so more than just for 100 MW) as backups, as well as having more energy capacity, as they believe they will continue to grow that data center.

7. He does believe there is some double booking going on on transformers and switchgears because of extra-long lead times.

8. Everyone is trying to reduce PUE, and water use effectiveness, but even after improving, they just use the same power to run more compute.

9. The problem is also liquid cooling, as it is expensive, and water availability in many regions is a problem.

10. Margins on equipment in the sector have gone from 4-6%, where they were 2-3 years ago, to 20-23% and in some cases even 40%. The data center builders know the margins are high, but they are fine with it because they just want to get it.

found on @AlphaSenseInc

[Alberto Rodriguez @_albertorod_]

You can’t lift a fridge with just your hands. Your whole body needs to conform to its shape, and bear the load between your arms and torso.

Here, @BostonDynamics' Atlas uses proprioception to manage the whole-body interaction and adapt to a shifting 100+ lb load. Enabling this type of high performance manipulation is exactly why we walked away from what was arguably the world’s best implementation of MPC for humanoids, and shifted entirely to RL without looking back.

This level of whole-body controls is a fundamental building block of physical intelligence and key to the value proposition of humanoids.

More technical details in:
Blog: https://t.co/oIRjVfh7jJ
Behind the scenes video: https://t.co/LgaImMAyhX

[Manisha Mishra @manishamishra24]

“Blue-collar jobs are safe from AI.”

Not for long.

This welding robot climbs vertical steel walls like Spider-Man.

While everyone is watching AI replace coders and office work…

industrial robots are quietly entering the physical world too.

Welding.
Grinding.
Inspection.
Rust removal.

The scary part?

Software automation was the easy phase.

Now machines are learning to work in the real world.

[Jacek @Jacek972]

@marcin_pod97229 @czaplinski_s @KWyszkowski Prowokuje? On należy do tych co nie uwierzyli, modlimy się za nich w Wielki Piątek.

[Natalie Wolchover @nattyover]

Bacteria move around using a molecular machine called the flagellar motor that rotates faster than the flywheel of a race car engine and switches directions in an instant. After 50 yrs, scientists have finally figured out how it works. “My lifelong quest is now fulfilled.” Link⤵️