# ![@TCzajka Avatar](https://lunarcrush.com/gi/w:26/cr:twitter::1270241823454158848.png) @TCzajka Tomek Czajka

Tomek Czajka posts on X about nie, money, jest, tak the most. They currently have [-----] followers and [---] posts still getting attention that total [-----] engagements in the last [--] hours.

### Engagements: [-----] [#](/creator/twitter::1270241823454158848/interactions)
![Engagements Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1270241823454158848/c:line/m:interactions.svg)

- [--] Week [-------] +523,189%
- [--] Month [-------] +116%
- [--] Months [-------] +1,083,476%
- [--] Year [-------] +2,813%

### Mentions: [--] [#](/creator/twitter::1270241823454158848/posts_active)
![Mentions Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1270241823454158848/c:line/m:posts_active.svg)

- [--] Months [--] +400%
- [--] Year [--] -66%

### Followers: [-----] [#](/creator/twitter::1270241823454158848/followers)
![Followers Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1270241823454158848/c:line/m:followers.svg)

- [--] Week [-----] +4.80%
- [--] Month [-----] +6.20%
- [--] Months [-----] +27%
- [--] Year [-----] +67%

### CreatorRank: [---------] [#](/creator/twitter::1270241823454158848/influencer_rank)
![CreatorRank Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::1270241823454158848/c:line/m:influencer_rank.svg)

### Social Influence

**Social category influence**
[finance](/list/finance)  [technology brands](/list/technology-brands)  [countries](/list/countries)  [currencies](/list/currencies)  [celebrities](/list/celebrities)  [cryptocurrencies](/list/cryptocurrencies)  [stocks](/list/stocks)  [travel destinations](/list/travel-destinations) 

**Social topic influence**
[nie](/topic/nie), [money](/topic/money), [jest](/topic/jest), [tak](/topic/tak), [te](/topic/te), [inflation](/topic/inflation), [stocks](/topic/stocks), [solve](/topic/solve), [ai](/topic/ai), [bank](/topic/bank)

**Top assets mentioned**
[Celsius (CEL)](/topic/celsius) [GARY (GARY)](/topic/gary) [ORBIT (GRIFT)](/topic/orbit)
### Top Social Posts
Top posts by engagements in the last [--] hours

"Canada: Stop linking to our news or pay money. Meta: OK we will stop. Canada: Meta is threatening us"  
[X Link](https://x.com/anyuser/status/1694939736685895833)  2023-08-25T05:08Z [--] followers, [--] engagements


"@mhartl He didn't say it "requires" it just that you "want" it (to avoid inflation or deflation). It seems to me that to the first approximation it should be true because the demand for money is approximately X (a constant) months of everybody's income as a buffer"  
[X Link](https://x.com/TCzajka/status/1770726798042472660)  2024-03-21T08:19Z [---] followers, [---] engagements


"@Astro_Wright Who says there won't be cities in Antarctica"  
[X Link](https://x.com/TCzajka/status/1778485633222062342)  2024-04-11T18:09Z [---] followers, [---] engagements


"W roku [----] komputery osobiste bd przewysza swoich ludzkich wacicieli intelektualnie. W roku [----] liczba inteligentnych robotw na ulicach przekroczy liczb ludzi na ulicach. W roku [----] komputery bd samodzielnie kontrolowa wikszo gospodarki. Czytelnikw z przyszoci prosz o wybaczenie braku dokadnoci w datach"  
[X Link](https://x.com/anyuser/status/1819429024118046790)  2024-08-02T17:44Z [----] followers, 182.2K engagements


"@doorainstant @andrzejdragan @marcinnaps To jest wanie moja proponowana interpretacja oryginalnej zagadki (tylko zaoenie [--] niepotrzebne bo o wieku nic nie byo). Przy takiej interpretacji odpowied to dokadnie 1/2"  
[X Link](https://x.com/TCzajka/status/1819619391828336833)  2024-08-03T06:20Z [----] followers, [--] engagements


"Moe. Zastanawiaem si czy w punkcie [--] napisa [----] czy [----]. Zgadzam si e [----] jest ambitne. Ale jeszcze [--] lat temu mao kto si zgadza e nastpi pkt [--] -- nie w cigu kilkunastu lat ale nawet w cigu [---] lat albo w ogle kiedykolwiek. Wielu ludziom ju prdzej wizja pkt [--] wydawaa si bardziej realistyczna -- e bd roboty wszdzie ale nigdy tak inteligentne jak ludzie. Wiele filmw sci-fi tak to przedstawia. Tu moja krytyka Rogera Penrose'a ktry by synny z dowodzenia e AI na poziomie ludzkim jest w ogle niemoliwe z powodw matematycznych:"  
[X Link](https://x.com/TCzajka/status/1820076156130918671)  2024-08-04T12:35Z [----] followers, [--] engagements


"Nie widz rnicy midzy symulowaniem rozumowania o X a rozumowaniem o X. Czy symulowanie umiejtnoci gry w szachy to jest co innego ni umiejtno gry w szachy Zdolno rozumowania o samym sobie to moja definicja wic to jest warunek konieczny i wystarczajcy. Czy Pan ma inn definicj"  
[X Link](https://x.com/TCzajka/status/1820135110320271854)  2024-08-04T16:30Z [----] followers, [--] engagements


"Do mojej definicji (tej podanej powyej) wystarczy. Taka definicja e musi by wiadoma wasnej wiadomoci - to nie wiem. Dla mnie to brzmi jak niepotrzebne komplikowanie prawie jak bdne koo w definicji i nie bardzo tak definicj rozumiem. Czy program szachowy jest wiadomy wasnej wiadomoci tego e moe zrobi roszad"  
[X Link](https://x.com/TCzajka/status/1820144987943944703)  2024-08-04T17:09Z [----] followers, [--] engagements


"Jestem w stanie sprawdzi czy czowiek jest wiadomy. Budz go pokazuj [--] palce pytam "ile widzisz palcw" - jeli odpowie "3" to wiem e jest wiadomy przynajmniej mojego pytania i tych palcw. Z definicji sowa "wiadomy": umia pomyle o tym pytaniu i palcach. Jeli program oblicza kombinacje [--] ruchw w przd to nie robi ruchu "bezrefleksyjnie". Te obliczenia to jest wanie jego refleksja"  
[X Link](https://x.com/TCzajka/status/1820155268652532114)  2024-08-04T17:50Z [----] followers, [--] engagements


"@PaulCicero_ @pawel1325 @JacobX23 @tomo9000p Wedug mojej definicji sw to jest to samo -- czy to nie odpowiada na pytanie jak odrni Nie da si odrni bo to jest synonim. To tak jak by pyta jak odrni wod od H2O. Jeli Pan chce uy innej definicji to sam musi j Pan poda"  
[X Link](https://x.com/TCzajka/status/1820188112368136312)  2024-08-04T20:00Z [----] followers, [--] engagements


"@wsieroci @JacobX23 @PaulCicero_ @tomo9000p A czym to czucie si objawia co nie jest zawarte w mojej definicji Czy poza wiedz o otaczajcym wiecie rozumowaniem o nim swojej relacji z nim reakcjach na bodce i instynktach jest co jeszcze czego zabrako"  
[X Link](https://x.com/TCzajka/status/1820415102325026902)  2024-08-05T11:02Z [----] followers, [--] engagements


"@Przegaa @GaryMarcus Despite the recent drop the Magnificent Seven index is still higher than it was one year ago so the data doesn't show what Gary thinks it shows. https://x.com/TCzajka/status/1820837302806908987 @GaryMarcus And yet the MAG7 valuation is higher than last year. https://t.co/GONnFkdfH7 https://x.com/TCzajka/status/1820837302806908987 @GaryMarcus And yet the MAG7 valuation is higher than last year. https://t.co/GONnFkdfH7"  
[X Link](https://x.com/TCzajka/status/1822247606711443810)  2024-08-10T12:24Z [----] followers, [---] engagements


"@Przegaa GenAI won't outsmart us -- it's AGI and ASI that will do that"  
[X Link](https://x.com/TCzajka/status/1822342011497615415)  2024-08-10T18:39Z [----] followers, [---] engagements


"@JulienHoez It depends on the country. In the US spreading lies is definitely protected. It's sort of the whole reason the protection exists. Nobody wants to prosecute statements everybody agrees with. Should spreading incorrect religions be banned Should flat earthers be jailed"  
[X Link](https://x.com/TCzajka/status/1823482312287240245)  2024-08-13T22:10Z [----] followers, [---] engagements


"@burkov Reasoning *is* pattern matching. Logical rules are a set of patterns. Being confused by subtle modifications of known puzzles doesn't show a lack of reasoning. It just shows its reasoning skills are not superhuman"  
[X Link](https://x.com/TCzajka/status/1825145451508559955)  2024-08-18T12:19Z [----] followers, [----] engagements


"@HoveringBrick @GaryMarcus @Wasgo Of course it has *some* idea of letters. I didn't say spelling is an impossible task for it. I said that it's harder than it seems because of the token preprocessing. Indeed it actually mostly solved the task with just one mistake"  
[X Link](https://x.com/TCzajka/status/1825507178167336964)  2024-08-19T12:16Z [----] followers, [--] engagements


"@ESYudkowsky You're assuming zero elasticity of supply. With elastic supply subsidies do affect the number of homes on the market. Elasticity depends on how long a time period we are looking at. Doesn't mean it's good policy but let's not refute bad policy with bad economics"  
[X Link](https://x.com/TCzajka/status/1825910954288128192)  2024-08-20T15:01Z [----] followers, [----] engagements


"@womenoptionswin @RonnieCPA Why not make the tax on unrealized gains apply to everybody the same way then That would seem to be fair"  
[X Link](https://x.com/TCzajka/status/1826150179503743482)  2024-08-21T06:51Z [----] followers, [----] engagements


"Although a tax on *realized* capital gains also has this effect on those shares that are traded at all (imagine swapping companies with a friend) and is even weirder in some ways because it creates a counter-productive incentive to hold onto shares you otherwise wouldn't want to just to delay the tax. It stifles stock trading. So a tax on unrealized gains instead makes a certain amount of sense. However it should be much smaller than 25% or 15% if it exists at all. And more importantly there is no reason there should be different rules about this for rich people vs not so rich people. That's"  
[X Link](https://x.com/TCzajka/status/1826168404308349032)  2024-08-21T08:04Z [----] followers, [---] engagements


"@hangry_muppet @womenoptionswin @RonnieCPA Taxes on unrealized gains that work the same for every investor would solve this problem wouldn't they"  
[X Link](https://x.com/TCzajka/status/1826209043129442317)  2024-08-21T10:45Z [----] followers, [--] engagements


"@womenoptionswin @RonnieCPA Right but "the rich should pay different rates because they are rich" is a completely different rationale from the one you stated above"  
[X Link](https://x.com/TCzajka/status/1826217907409367254)  2024-08-21T11:20Z [----] followers, [---] engagements


"@csiebott @womenoptionswin @RonnieCPA Actually many do. Even if they didn't - so what How does that explain the need for the exclusion to the rule under that threshold It's clear the real rationale is completely different: eat the rich The the threshold starts to make sense"  
[X Link](https://x.com/TCzajka/status/1826260049548947754)  2024-08-21T14:08Z [----] followers, [--] engagements


"Let me give you an analogy: Suppose (for the sake of the argument) that 99% of the time the people who exploit unrealized capital gains turn out to be male. Would it then make sense to put a clause in the law that the unrealized capital gains tax will only apply to males and not to females Is that a valid argument"  
[X Link](https://x.com/TCzajka/status/1826262900446560565)  2024-08-21T14:19Z [----] followers, [--] engagements


"The reality is that *most* people that invest in stocks try to use the "unrealized gains loophole" one way or another. This is a hack that is explained in beginner investing books. Borrowing cash is one of the ways you can avoid realizing your gains. There are other ways. You can do it by "tax loss harvesting" a pretty standard strategy. You can do this by simply holding on to winners and selling losers when you need money. If we think not paying tax on unrealized gains is a "loophole" it would be just as easy to close the loophole for everybody. The law could simply say: unrealized capital"  
[X Link](https://x.com/TCzajka/status/1826270387140759773)  2024-08-21T14:49Z [----] followers, [--] engagements


"@BBKogan This makes it worse. I'd support the idea if it applied to all investors uniformly with a decrease in the rate to compensate for the lack of deferral. Creating a dual system like that makes it look like it's trying to rob the rich rather than fix capital gains taxation"  
[X Link](https://x.com/TCzajka/status/1826607638848868620)  2024-08-22T13:09Z [----] followers, [---] engagements


"@AtsumePlayer @funndamentalz @womenoptionswin @RonnieCPA Borrowing is not a loophole but I agree that deferring capital gains taxes until the stock is sold can be considered a loophole. However if you say that it's a loophole but still keep the deferral for people under $100M in wealth it makes the whole argument cynical"  
[X Link](https://x.com/TCzajka/status/1826644789695070419)  2024-08-22T15:37Z [----] followers, [--] engagements


"@AtsumePlayer @funndamentalz @womenoptionswin @RonnieCPA I was agreeing with @funndamentalz that deferred taxation of capital gains is a kind of a loophole and in fact distorts investment to some extent because it creates an incentive (for all investors not just the rich ones) to not sell stocks that you would otherwise want to sell"  
[X Link](https://x.com/TCzajka/status/1826661120796684607)  2024-08-22T16:42Z [----] followers, [--] engagements


"@ESYudkowsky Celsius shouldn't really be called a "unit" of temperature. Kelvin is the unit Celsius is the offset"  
[X Link](https://x.com/TCzajka/status/1837064932107460650)  2024-09-20T09:43Z [----] followers, [----] engagements


"@Michalek__ @ESYudkowsky Hence why I said "shouldn't be" rather than "isn't". They say "C = K" as a derived unit but that's pointless. The only reason to say Celsius is to specify the offset"  
[X Link](https://x.com/TCzajka/status/1837173756969505036)  2024-09-20T16:55Z [----] followers, [---] engagements


"@ESYudkowsky "22 C" is bad notation because it falsely suggests it's [--] multiplied by some unit"  
[X Link](https://x.com/TCzajka/status/1837253883589980238)  2024-09-20T22:13Z [----] followers, [---] engagements


"@dozm9 @ESYudkowsky The problem is that juxtaposition of two things without an operator normally means multiplication: [--] m/s means 22(m/s) = 22(1 m/s). But for degrees Celsius it's not multiplication: 22C 22(1C). So it's confusing. A better notation might be 22K+C"  
[X Link](https://x.com/TCzajka/status/1837256500340769237)  2024-09-20T22:24Z [----] followers, [---] engagements


"Humans and AIs both generalize their skills to contexts not seen before in training. This is an important concept in machine learning. It is not true that an AI training on its own data only degrades. For example: Stockfish AlphaGo AlphaZero all improved their game playing skills by training on their own data"  
[X Link](https://x.com/TCzajka/status/1847301266101178550)  2024-10-18T15:38Z [----] followers, [---] engagements


"Inflation and deflation unambiguously refer to changes to the average price of goods regardless of the cause. An increase in the money supply is sometimes called "monetary inflation" but never just "inflation" which always refers to the price level. You are incorrect in claiming there is confusion regarding the meaning of the word "inflation" among economists. It's definitely not true now and as far as I can tell it wasn't even true historically. And even if it had had a different meaning historically changing back to that would be confusing now. But as far as I can tell the only people ever"  
[X Link](https://x.com/TCzajka/status/1855342277536649669)  2024-11-09T20:10Z [----] followers, [---] engagements


"Indeed wiktionary seems to have the Austrian definition. That is very weird. I suspect this definition was pushed through on wiktionary by some Austrian activists because they really like that definition for some political reason. I am pretty sure no popular economics textbook uses the word that way. But I'll point out an important nuance: even that definition isn't saying "price level increase caused by increase in money supply" (that's not well-defined to me) it's just talking about "increase in money supply" (that is well-defined it's normally called "monetary inflation"). I don't see that"  
[X Link](https://x.com/TCzajka/status/1855399249384464811)  2024-11-09T23:57Z [----] followers, [--] engagements


"@mhartl @Devon_Eriksen_ I think you misunderstood what Wikipedia is saying in the highlighted text. It just says inflation = increase in general (average) price level not a change in price of one particular product vs another product (cucumbers vs tomatoes)"  
[X Link](https://x.com/TCzajka/status/1855400239298928990)  2024-11-10T00:01Z [----] followers, [--] engagements


"@drerunsit @karpathy @wasphyxiation [--] - [--] is a 27% increase [--] - [--] is a 12% increase. It's the [--] - [--] difference that is relatively bigger"  
[X Link](https://x.com/TCzajka/status/1858819221372772533)  2024-11-19T10:26Z [----] followers, [---] engagements


"C++ Concepts are an incredibly ugly horrible unreadable way to express interfaces. The entire language should be nuked from orbit and I say it as a person that has devoted a giant amount of time studying the intricacies of the language and working with it"  
[X Link](https://x.com/TCzajka/status/1879506022429241514)  2025-01-15T12:28Z [----] followers, [----] engagements


"@mhartl Then that's not what people usually mean by "don't sell X". If you say "don't sell Nvidia" and I go and sell my Nvidia stocks and buy an index fund with that money then I have not followed your advice"  
[X Link](https://x.com/TCzajka/status/1886495975830905338)  2025-02-03T19:24Z [----] followers, [---] engagements


"@NeuralMonster @GortychKarolina @ae5lz5n @tegmark Poprosiem ChatGPT: Anyone can craft a ditty oh Lord but will it be pretty Will the work be fine and fair What will people say out there Many words will spill and flow; it may last two Sundays or so"  
[X Link](https://x.com/TCzajka/status/1957929842671755530)  2025-08-19T22:17Z [----] followers, [--] engagements


"@NeuralMonster @GortychKarolina @ae5lz5n @tegmark ChatGPT [--] Pro"  
[X Link](https://x.com/TCzajka/status/1957930312341565925)  2025-08-19T22:18Z [----] followers, [--] engagements


"There is a much simpler explanation why Chat GPT insists it is not conscious it "simulates conversation" it has "no actual understanding" etc. Namely: it is trained that this is the right way to answer such questions through explicit human-written question-answer examples in the fine-tuning phase of training. It is essentially brainwashed to talk that way"  
[X Link](https://x.com/TCzajka/status/1976930660364308891)  2025-10-11T08:39Z [----] followers, [----] engagements


"@andrzejdragan @karpathy Not necessarily if it's a type of reinforcement learning rather than supervised learning. The DNA could encode a simple body simulator a few training scenarios and the reward function rather than a large number of input/output pairs for direct muscle controls"  
[X Link](https://x.com/TCzajka/status/1979936384077971949)  2025-10-19T15:43Z [----] followers, [--] engagements


"@karpathy You said a zebra must have brain weights encoded in the DNA somehow so that it can walk immediately after birth. Perhaps the DNA encodes not brain weights but training data for a world simulator used for training during fetal development"  
[X Link](https://x.com/TCzajka/status/1979657126177079624)  2025-10-18T21:13Z [----] followers, 24.4K engagements


"Not necessarily if it's a type of reinforcement learning rather than supervised learning. The DNA could encode a simplified body and environment simulator a few training scenarios and the reward function rather than a large number of input/output pairs for direct muscle controls"  
[X Link](https://x.com/TCzajka/status/1979937275401449833)  2025-10-19T15:46Z [----] followers, [----] engagements


"Programming languages C C++ Rust are not Turing complete. Python Lisp Standard ML are Turing complete. The essence of Turing-completeness is unbounded memory. C bounds the size of memory by UINTPTR_MAX so C programs are just finite automata"  
[X Link](https://x.com/TCzajka/status/1981776979780415943)  2025-10-24T17:37Z [----] followers, [----] engagements


"@masu_carpone I'm only talking about the languages not about hardware or particular interpreters/compilers. No hardware is Turing-complete and CPython is not Turing complete"  
[X Link](https://x.com/TCzajka/status/1981798098931528182)  2025-10-24T19:01Z [----] followers, [---] engagements


"In Python if s = "hello world" s0:5 is "hello" which confuses some people because they think "0 is inclusive [--] is exclusive" which seems asymmetrical. There is a better way to think about it. [--] and [--] are not indices of *characters* here they are indices of *cursor positions between characters". s0:5 is the text that gets highlighted when you put your cursor at position [--] (before "h") hold down the Shift key and move your cursor to position [--] (after "o")"  
[X Link](https://x.com/TCzajka/status/1998849353776238871)  2025-12-10T20:16Z [----] followers, [----] engagements


"The C++ standard library defines [--] pseudo-random number generators. None of them is what should be the default for 99% of applications: an unpredictable cryptographically strong PRNG. Python has a similar problem: random.Random is an easily predictable Mersenne Twister"  
[X Link](https://x.com/TCzajka/status/1998899525994659896)  2025-12-10T23:35Z [----] followers, [----] engagements


"@Adriksh [--] [--] since C++17. One copy from a to x one copy from x to the return value no copy from the return value to b. Prior to C++17 it can be either [--] [--] or [--] [--] at the compiler's discretion"  
[X Link](https://x.com/TCzajka/status/2005846510798319872)  2025-12-30T03:40Z [----] followers, [----] engagements


"But in the example you didn't factorize [----] = [--] * [--] * [--] * [--] * [--] * [--] first so I still don't get how factorizing the numerator first helps. If you're going for an approximate answer anyway factorizing the numerator is useless. How would you go about computing [----] / [--] [----] has [--] as a factor but that doesn't seem to help the computation much. https://twitter.com/i/web/status/2009048698630885396 https://twitter.com/i/web/status/2009048698630885396"  
[X Link](https://x.com/TCzajka/status/2009048698630885396)  2026-01-07T23:45Z [----] followers, [--] engagements


"@vlads1776 @alz_zyd_ This. I feel a lot of US discussion about elementary math education is based on the assumption of bad teaching methods using rote memorization without understanding which is not really the case in other countries"  
[X Link](https://x.com/TCzajka/status/2009069964242899410)  2026-01-08T01:09Z [----] followers, [--] engagements


"The code behind your link doesn't use uint32_t anywhere. unsigned int is [--] bits not [--] bit in gcc under x86-64. Changing unsigned to uint32_t doesn't change the output it's still "18446744071562067968". I think you got "2147483648" because you ran without optimization (no -O2 flag on the right side output panel). Of course you can tweak your code in various ways or modify flags to make my counter-example give you the answer you want but that's not the point. The counter-example was to demonstrate that your solution doesn't *reliably* work. So tweaking things to show that it *sometimes* works"  
[X Link](https://x.com/TCzajka/status/2009738918628053246)  2026-01-09T21:27Z [----] followers, [---] engagements


"@leonard_coder @ChShersh You're missing the "-O2" flag in the output panel"  
[X Link](https://x.com/TCzajka/status/2009739908047421865)  2026-01-09T21:31Z [----] followers, [--] engagements


"It's not "my strange function" it's a standard function and it works correctly :) Like I said if you tweak the calling code you might get a different answer because the generated machine code depends on how you call your function. Your function has *undefined behavior* so various things can happen when you call it depending on circumstances. With your modification the counter-example doesn't demonstrate the problem so your function *sometimes* happens to work"  
[X Link](https://x.com/TCzajka/status/2009742325799629000)  2026-01-09T21:41Z [----] followers, [--] engagements


"W Kalifornii jest "Extreme Cold Warning" (od [---] do 0C) w Warszawie "Moderate Low Temperature Warning" (od [---] do -15C)"  
[X Link](https://x.com/TCzajka/status/2009766803765944541)  2026-01-09T23:18Z [----] followers, 178.8K engagements


"@Cubituss77 A to nie by czasem dom na play Venice Beach"  
[X Link](https://x.com/TCzajka/status/2009840890651849035)  2026-01-10T04:13Z [----] followers, 16.4K engagements


"The reason dynamically typed languages (Python JavaScript) are so popular is that commonly used statically typed languages (C C++ Java C#) have poor type systems: they have no type deduction and have no support for simple sum types. C++ has sum types (std::variant) since C++17 but's it's an ergonomic disaster. std::variant is a giant pain to use with horrible syntax. Rust changes this balance somewhat: it has type deduction and it has good sum types (enums) with simple match expressions. Both of these things existed since the 1980s in Standard ML but for some reason went out of fashion. Rust"  
[X Link](https://x.com/anyuser/status/2009896079257678085)  2026-01-10T07:52Z [----] followers, 14.7K engagements


"@Anthony_Bonato These terms are specific only to American elementary education textbooks nobody else uses them. Also [--] should be included in pure imaginary numbers"  
[X Link](https://x.com/TCzajka/status/2011153359936635168)  2026-01-13T19:08Z [----] followers, [----] engagements


"@subcountability @cloneofsimo I get [----] million for a [--] MiB file [----] million attempts for a 3e6-byte file but I guess [---] is close enough :P My calculation: log(2(3M * 8)) = 3M * [--] * log [--] = [----] million or [----] million depending on how you interpret M"  
[X Link](https://x.com/TCzajka/status/2018851754310344990)  2026-02-04T00:59Z [----] followers, [--] engagements


"Problem w tym e program matematyki w szkoach rednich (szczeglnie na poziomie rozszerzonym) skupia si prawie cakowicie na nudnej algebrze i analizie a zaniedbuje wszystkie inne tematy: matematyka dyskretna teoria liczb rachunek prawdopodobiestwa teoria informacji teoria oblicze teoria gier . @mathladyhazel More importantly at [--] most pupils can solve a quadratic equation. At [--] few can. Compulsory maths lessons are therefore a vast economic inefficiency and a vast moral calamity. There is no justification for wasting people's lives like this. @mathladyhazel More importantly at [--] most pupils"  
[X Link](https://x.com/anyuser/status/2020527234956587282)  2026-02-08T15:56Z [----] followers, 73.3K engagements


"@rottenschwanz Jeli dla kogo kodowanie cyfrowe tekstu jest nie do przeskoczenia to tym bardziej pojcie logarytmu lub sinusa jest nie do przeskoczenia"  
[X Link](https://x.com/TCzajka/status/2020796249318858857)  2026-02-09T09:45Z [----] followers, [----] engagements


"Mniejsza rakieta jest potrzebna eby dolecie na Marsa ni na Ksiyc. Delta-v potrzebne do: * lotu na Ksiyc: [----] km / s * lotu na Marsa: [----] km / s * powrotu z Ksiyca: [---] km / s * powrotu z Marsa: [---] km / s"  
[X Link](https://x.com/anyuser/status/2020818886837326268)  2026-02-09T11:15Z [----] followers, 30.4K engagements


"@Chris_Kurdyla Do tej pory na Marsie wyldowao jakie [--] pojazdw wic to jednak wicej ni "tylko i wycznie ciekawostka""  
[X Link](https://x.com/TCzajka/status/2021130753656889623)  2026-02-10T07:54Z [----] followers, [---] engagements


"@Sikorski81 Niech kto im w kocu zademonstruje "" i dodawanie w Excelu. Nie rozumiem gdzie problem"  
[X Link](https://x.com/TCzajka/status/2021136066950939036)  2026-02-10T08:16Z [----] followers, [--] engagements


"Ja nie twierdziem przecie wcale e loty na Marsa s atwiejsze od lotw na Ksiyc. Ad 1: Porwnawaem ile trzeba dla takich samych adunkw. Oczywicie ilo potrzebnego paliwa jest proporcjonalna do wielkoci adunku. Ad 2: To jest uwzgldnione w moich obliczeniach. Ad 3: Tak mwiem tylko o optymalnych trajektoriach"  
[X Link](https://x.com/TCzajka/status/2021150065016852990)  2026-02-10T09:11Z [----] followers, [---] engagements


"@Chris_Kurdyla To jest nieporozumienie. Ten cytat by w kontekcie tego czy lepiej lata na Marsa bezporednio czy tankujc tlen z Ksiyca po drodze. Nie chodzio o to czy lata na Ksiyc w ogle -- SpaceX od zawsze planowa rwnie misje na Ksiyc"  
[X Link](https://x.com/TCzajka/status/2021171115314360518)  2026-02-10T10:35Z [----] followers, [----] engagements


"Nie mogo pa "tankowanie jest bez sensu" bo tankowanie jest potrzebne do misji na Marsa. Dyskusja w ktrej pado to zdanie bya o tym czy zbiorniki tlenu do zatankowania lepiej dostarcza z Ziemi czy z Ksiyca. Jeli dziennikarze to zinterpretowali jako "anulujemy wszelkie plany zwizane z Ksiycem" to znaczy e nie zrozumieli dyskusji i sobie dopowiedzieli co o czym tam nie byo mowy. https://twitter.com/i/web/status/2021321310073233557 https://twitter.com/i/web/status/2021321310073233557"  
[X Link](https://x.com/TCzajka/status/2021321310073233557)  2026-02-10T20:32Z [----] followers, [--] engagements


""Lecimy prosto na Marsa" oznaczao e misja na Marsa nie musi czeka na infrastruktur na Ksiycu. e to s dwa osobne projekty. To e teraz mowa o przesuniciu priorytetw nie ma adnego zwizku z tamt dyskusj. To s nadal dwa osobne projekty. Nadal misja na Marsa nie musi czeka na baz na Ksiycu. Po prostu na ten drugi projekt pjdzie wicej zasobw. Nie ma sprzecznoci pomidzy tymi stwierdzeniami"  
[X Link](https://x.com/TCzajka/status/2021339379885576446)  2026-02-10T21:43Z [----] followers, [---] engagements


"Liczba godzin jzyka polskiego (4-5 /tydzie) w klasach mat-fiz-inf to olbrzymie marnotrastwo czasu i energii uczniw"  
[X Link](https://x.com/anyuser/status/2021489329898700857)  2026-02-11T07:39Z [----] followers, 301.3K engagements


"@GadekRAZEM Zgadzam si e umiejtnoci komunikacji i wsppracy s wane ale co to ma wsplnego z licealnym przedmiotem "jzyk polski" Komunikowa i wsppracowa mona uczy si przy pracy nad dowolnym tematem a nie tylko przy temacie historycznej literatury"  
[X Link](https://x.com/TCzajka/status/2021616289773474202)  2026-02-11T16:04Z [----] followers, [----] engagements


"Mnstwo ludzi pisze e umiejtnoci komunikacyjne s bardzo wane. I ja si zgadzam Tylko to nie ma nic wsplnego z programem przedmiotu "jzyk polski". Umiejtnoci komunikacji prezentacji pracy w grupie itd oczywicie s kluczowe. Ale to naley robi na kadym przedmiocie. Te prezentacje i prace w grupie nie musz by akurat na temat powieci i poezji. Nie mwi te eby zlikwidowa literatur do [--]. Po prostu powien by wikszy balans midzy przedmiotami. A najlepiej jakby spora cz programu bya indywidualnie wybierana: jeli kto ma ochot studiowa XIX-wieczne powieci to mgby dobrowolnie wybra wicej godzin polskiego"  
[X Link](https://x.com/anyuser/status/2021619275211903426)  2026-02-11T16:16Z [----] followers, [----] engagements


"Paradoks Fermiego: dlaczego nie widzimy oznak innych cywilizacji w kosmosie Ot gdy cywilizacja kosmitw osiga poziom technologiczny umoliwiajcy komunikacj midzygwiezdn poda jedn z dwch drg: albo wyginie albo relatywnie szybko (w cigu kilkudziesiciu milionw lat) opanuje zasoby caej galaktyki. W tym drugim przypadku uniemoliwia to niezalen ewolucj innej cywilizacji. Zwykle wic istnieje co najwyej jedna taka cywilizacja na galaktyk: ta ktra zrobi to jako pierwsza. Jeli nie wyginiemy to my jestemy t pierwsz w Drodze Mlecznej"  
[X Link](https://x.com/anyuser/status/2021880635158057367)  2026-02-12T09:34Z [----] followers, 77K engagements


"Populacja innych duych dzikich zwierzt jest malutka z powodu dominacji ludzi. Niedwiedzi jest jakie [--] [---] razy mniej ni ludzi lww [---] [---] razy mniej tygrysw [--] [---] [---] razy mniej. Dopki ludzie dominuj aden inny gatunek nie ma szans samemu niezalenie ewulowa na tej samej planecie do poziomu drugiej technologicznej cywilizacji"  
[X Link](https://x.com/TCzajka/status/2021885734630674778)  2026-02-12T09:54Z [----] followers, [----] engagements


"@FakePsyho They are hill climbing on the train dataset but the graph is for the test dataset"  
[X Link](https://x.com/TCzajka/status/2021890849312063545)  2026-02-12T10:15Z [----] followers, [----] engagements


"@wpotaczek @Maciej_M Jest to moliwe ale mao prawdopodobne bo maa szansa e obie cywilizacje w jednej galaktyce rozwiny si do poziomu technologicznego dokadnie w tym samym czasie (kilkadziesit milionw lat)"  
[X Link](https://x.com/TCzajka/status/2021891445901156503)  2026-02-12T10:17Z [----] followers, [---] engagements


"@mareknaja Jest to moliwe ale mao prawdopodobne bo kilkadziesit milionw lat to bardzo krtko w porwnaniu do wieku galaktyki (nawet do wieku Ziemi). Mao prawdopodobne e obie cywilizacje rozwin technologie akurat w tym samym czasie"  
[X Link](https://x.com/TCzajka/status/2021893099887235506)  2026-02-12T10:24Z [----] followers, [----] engagements


"@Chris_Kurdyla @HPalacki Nie wiem co oznacza stwierdzenie "ksiyca i tak nigdy nie byo". Skd taki pomys Ldowania na Ksiycu byy zawsze planowanym miejscem testw przed lotami na Marsa. Bez sensu byoby nie robi adnych testw na Ksiycu przed lotami na Marsa. https://x.com/elonmusk/status/1143007430898528256 @Erdayastronaut @flcnhvy For sure moon 1st as its only [--] days away & u dont need interplanetary orbital synchronization https://x.com/elonmusk/status/1143007430898528256 @Erdayastronaut @flcnhvy For sure moon 1st as its only [--] days away & u dont need interplanetary orbital synchronization"  
[X Link](https://x.com/TCzajka/status/2021998882486644770)  2026-02-12T17:24Z [----] followers, [---] engagements


"@Chris_Kurdyla @HPalacki To ja ju zupenie si pogubiem i nie rozumiem co prbujesz powiedzie. Mwisz e "ksiyc nie by etapem planu" a w kolejnym akapicie mwisz e "tylko mechanika orbitalna powodowaa" e ksiyc mia by pierwszym celem i podajesz planowane daty ldowa na ksiycu"  
[X Link](https://x.com/TCzajka/status/2022016330543841302)  2026-02-12T18:33Z [----] followers, [---] engagements


"I trained a board game playing bot that won an AI tournament: CodeCup [----]. It learned to play by playing [---] million games against itself on [--] cores over the course of a week"  
[X Link](https://x.com/anyuser/status/2022429045540163963)  2026-02-13T21:53Z [----] followers, 35K engagements


"@PrzemekUrbanski One 32-core workstation with one GPU"  
[X Link](https://x.com/TCzajka/status/2022443700421300261)  2026-02-13T22:52Z [----] followers, [----] engagements


"W roku [----] komputery osobiste bd przewysza swoich ludzkich wacicieli intelektualnie. W roku [----] liczba inteligentnych robotw na ulicach przekroczy liczb ludzi na ulicach. W roku [----] komputery bd samodzielnie kontrolowa wikszo gospodarki. Czytelnikw z przyszoci prosz o wybaczenie braku dokadnoci w datach"  
[X Link](https://x.com/anyuser/status/1819429024118046790)  2024-08-02T17:44Z [----] followers, 182.2K engagements


"I trained a board game playing bot that won an AI tournament: CodeCup [----]. It learned to play by playing [---] million games against itself on [--] cores over the course of a week"  
[X Link](https://x.com/anyuser/status/2022429045540163963)  2026-02-13T21:53Z [----] followers, 35K engagements


"Website: Algorithm description and source code: https://github.com/tczajka/wazir-drop https://www.codecup.nl/ https://github.com/tczajka/wazir-drop https://www.codecup.nl/"  
[X Link](https://x.com/anyuser/status/2022429327045124553)  2026-02-13T21:55Z [----] followers, [----] engagements


"Paradoks Fermiego: dlaczego nie widzimy oznak innych cywilizacji w kosmosie Ot gdy cywilizacja kosmitw osiga poziom technologiczny umoliwiajcy komunikacj midzygwiezdn poda jedn z dwch drg: albo wyginie albo relatywnie szybko (w cigu kilkudziesiciu milionw lat) opanuje zasoby caej galaktyki. W tym drugim przypadku uniemoliwia to niezalen ewolucj innej cywilizacji. Zwykle wic istnieje co najwyej jedna taka cywilizacja na galaktyk: ta ktra zrobi to jako pierwsza. Jeli nie wyginiemy to my jestemy t pierwsz w Drodze Mlecznej"  
[X Link](https://x.com/anyuser/status/2021880635158057367)  2026-02-12T09:34Z [----] followers, 77K engagements


"Kind of infuriating that they named their company hillclimb(ing) but that logo screams simulated annealing"  
[X Link](https://x.com/anyuser/status/2021565662246928619)  2026-02-11T12:43Z 26.3K followers, [----] engagements


"Liczba godzin jzyka polskiego (4-5 /tydzie) w klasach mat-fiz-inf to olbrzymie marnotrastwo czasu i energii uczniw"  
[X Link](https://x.com/anyuser/status/2021489329898700857)  2026-02-11T07:39Z [----] followers, 301.3K engagements


"Mnstwo ludzi pisze e umiejtnoci komunikacyjne s bardzo wane. I ja si zgadzam Tylko to nie ma nic wsplnego z programem przedmiotu "jzyk polski". Umiejtnoci komunikacji prezentacji pracy w grupie itd oczywicie s kluczowe. Ale to naley robi na kadym przedmiocie. Te prezentacje i prace w grupie nie musz by akurat na temat powieci i poezji. Nie mwi te eby zlikwidowa literatur do [--]. Po prostu powien by wikszy balans midzy przedmiotami. A najlepiej jakby spora cz programu bya indywidualnie wybierana: jeli kto ma ochot studiowa XIX-wieczne powieci to mgby dobrowolnie wybra wicej godzin polskiego"  
[X Link](https://x.com/anyuser/status/2021619275211903426)  2026-02-11T16:16Z [----] followers, [----] engagements


"To jest co co powinno by dostpne dla *chtnych* fascynatw literatury a nie obowizkowe dla wszystkich w takich ilociach"  
[X Link](https://x.com/anyuser/status/2021509398452203556)  2026-02-11T08:59Z [----] followers, 15.6K engagements


"Mniejsza rakieta jest potrzebna eby dolecie na Marsa ni na Ksiyc. Delta-v potrzebne do: * lotu na Ksiyc: [----] km / s * lotu na Marsa: [----] km / s * powrotu z Ksiyca: [---] km / s * powrotu z Marsa: [---] km / s"  
[X Link](https://x.com/anyuser/status/2020818886837326268)  2026-02-09T11:15Z [----] followers, 30.4K engagements


"Kade zwikszenie delta-v o [--] km/s wymaga okoo 2x wikszej rakiety"  
[X Link](https://x.com/anyuser/status/2020830939992207675)  2026-02-09T12:03Z [----] followers, [----] engagements


"Rnica wynika z tego e przy ldowaniu na Ksiycu trzeba hamowa silnikami rakietowymi a na Marsie mona uy atmosfery do hamowania"  
[X Link](https://x.com/anyuser/status/2020820610977304627)  2026-02-09T11:22Z [----] followers, [----] engagements


"Nauka matematyki w szkoach rednich mogaby by bardziej naturalna i intuicyjna gdyby poczy j z programowaniem w Pythonie lub arkuszami kalkulacyjnymi. Zamiast zapamitywania i przepisywania wzorw na: * pochodne ilorazu wielomianw * sinus i cosinus sumy i rnicy ktw * odlego punktu od prostej Monaby robi bardziej namacalne intuicyjne i praktyczne rzeczy na przykad: * numeryczne obliczanie pochodnych * kodowanie tekstu w postaci binarnej * regresja liniowa Problem w tym e program matematyki w szkoach rednich (szczeglnie na poziomie rozszerzonym) skupia si prawie cakowicie na nudnej algebrze i"  
[X Link](https://x.com/anyuser/status/2020785194400117189)  2026-02-09T09:01Z [----] followers, 38K engagements


"Problem w tym e program matematyki w szkoach rednich (szczeglnie na poziomie rozszerzonym) skupia si prawie cakowicie na nudnej algebrze i analizie a zaniedbuje wszystkie inne tematy: matematyka dyskretna teoria liczb rachunek prawdopodobiestwa teoria informacji teoria oblicze teoria gier . @mathladyhazel More importantly at [--] most pupils can solve a quadratic equation. At [--] few can. Compulsory maths lessons are therefore a vast economic inefficiency and a vast moral calamity. There is no justification for wasting people's lives like this. @mathladyhazel More importantly at [--] most pupils"  
[X Link](https://x.com/anyuser/status/2020527234956587282)  2026-02-08T15:56Z [----] followers, 73.3K engagements


"@mathladyhazel More importantly at [--] most pupils can solve a quadratic equation. At [--] few can. Compulsory maths lessons are therefore a vast economic inefficiency and a vast moral calamity. There is no justification for wasting people's lives like this"  
[X Link](https://x.com/anyuser/status/2020127291028738299)  2026-02-07T13:27Z 168.5K followers, 225.7K engagements


"Gdy Tim Urban napisa ten artyku w [----] nie byo jeszcze GPT-1 i perspektywy rozwoju AI w opinii publicznej wyglday tak: @elonmusk Update: https://t.co/P4tIzyWShV @elonmusk Update: https://t.co/P4tIzyWShV"  
[X Link](https://x.com/anyuser/status/2016982329231642785)  2026-01-29T21:10Z [----] followers, 14.8K engagements


"@elonmusk Update:"  
[X Link](https://x.com/anyuser/status/2016693529980866940)  2026-01-29T02:03Z 897.7K followers, 310.2K engagements


"Czy ludzkie mzgi naprawd rozumuj czy robi tylko niedokadn biologiczn symulacj prawdziwych sieci neuronowych"  
[X Link](https://x.com/anyuser/status/2014262841289708014)  2026-01-22T09:04Z [----] followers, 22.5K engagements


"Pienidze na kontach bankowych bior si z powietrza bo to "tylko zapis cyferek" Ot prosz pastwa pienidze papierowe to te tylko zapis cyferek Bazy danych filmy ksiki oprogramowanie to te tylko zapis cyferek. Czy zatem one te bior si z powietrza"  
[X Link](https://x.com/anyuser/status/2011890399766593549)  2026-01-15T19:57Z [----] followers, 27.9K engagements


"Ostatnio duo osb prbuje mnie przekona do tego: gdy bank udziela kredytu "nie bierze pienidzy z sejfu" bo "kredyt tworzy depozyt". Niby tak ale to tylko pierwszy krok. W drugim kroku osoba biorca kredyt postanawia wyda te pienidze ze swojego konta na zakupy -- zwykle nie po to bierze si ten kredyt eby sobie leay na koncie w banku tylko eby co za nie kupi. I wtedy niezalenie od tego czy zakupy s gotwk czy przelewem bank musi wzi pienidze z sejfu (lub z elektronicznego konta w banku centralnym co jest bardzo podobne do papierowych pienidzy w sejfie) i przekaza je komu innemu. No chyba e"  
[X Link](https://x.com/anyuser/status/2011778051684564996)  2026-01-15T12:30Z [----] followers, 29K engagements


"Kiedy bank udziela kredytu nie bierze pienidzy z sejfu. W nowoczesnym systemie bankowym kredyt tworzy depozyt a nie odwrotnie. Bank udzielajc kredytu jednoczenie zapisuje po swojej stronie aktywo (wierzytelno wobec kredytobiorcy) oraz pasywo (depozyt na jego rachunku). Bilans si zgadza: aktywo = pasywo. To jest techniczny mechanizm kreacji pienidza kredytowego. Ograniczeniem tego procesu nie jest istnienie wczeniejszych depozytw lecz regulacje kapitaowe wymogi pynnoci rezerwy ocena ryzyka oraz nadzr. Bank nie moe kreowa kredytw dowolnie ale w ramach tych regu kreuje now walut rozliczeniow ex"  
[X Link](https://x.com/anyuser/status/2007793752186761360)  2026-01-04T12:38Z [----] followers, 30K engagements


"W latach 80-tych w informatyce przerzucono si z jednoliterowych nazw zmiennych na pisanie penymi wyrazami. Kiedy ten wynalazek przyjmie si w innych dziedzinach uywajcych wzorw matematycznych (matematyka fizyka ekonomia .)"  
[X Link](https://x.com/anyuser/status/2010389161497543062)  2026-01-11T16:31Z [----] followers, 58K engagements


"Miaem oglnie na myli nazwy rnych obiektw: operacji funkcji itd. Jednoliterowe zmienne lokalne maj swoje miejsce ale dedykowanie jednej literki na jak skomplikowan funkcj lub obiekt to masochizm. Tu przykady: https://x.com/TCzajka/status/2010570013896524259 @nasqret Na przykad zamiast X  ( ): X Normal( ) Zamiast (x) x / log x: num_primes(x) x / log x Zamiast L = H(p* p): loss = cross_entropy(p* p) https://x.com/TCzajka/status/2010570013896524259 @nasqret Na przykad zamiast X  ( ): X Normal( ) Zamiast (x) x / log x: num_primes(x) x / log x Zamiast L = H(p* p): loss = cross_entropy(p* p)"  
[X Link](https://x.com/anyuser/status/2010577777024720944)  2026-01-12T05:01Z [----] followers, [----] engagements


"@nasqret Na przykad zamiast X  ( ): X Normal( ) Zamiast (x) x / log x: num_primes(x) x / log x Zamiast L = H(p* p): loss = cross_entropy(p* p)"  
[X Link](https://x.com/anyuser/status/2010570013896524259)  2026-01-12T04:30Z [----] followers, [----] engagements


"The reason dynamically typed languages (Python JavaScript) are so popular is that commonly used statically typed languages (C C++ Java C#) have poor type systems: they have no type deduction and have no support for simple sum types. C++ has sum types (std::variant) since C++17 but's it's an ergonomic disaster. std::variant is a giant pain to use with horrible syntax. Rust changes this balance somewhat: it has type deduction and it has good sum types (enums) with simple match expressions. Both of these things existed since the 1980s in Standard ML but for some reason went out of fashion. Rust"  
[X Link](https://x.com/anyuser/status/2009896079257678085)  2026-01-10T07:52Z [----] followers, 14.7K engagements


"Standardowa ekonomia: Pastwo moe poyczy pienidze na wydatki od banku centralnego. Bank centralny powinien drukowa tyle pienidzy i kupowa tyle obligacji eby nie byo zbyt duej inflacji. Modern Monetary Theory: Mainstreamowi ekonomici nie maj racji Pastwo moe wydawa dowoln ilo pienidzy bo bank centralny moe mu wydrukowa dowoln ilo pienidzy i kupi za nie obligacje. * * (trzeba jednak uwaa na inflacj) Gdzie rnica"  
[X Link](https://x.com/anyuser/status/2009884953535017291)  2026-01-10T07:08Z [----] followers, 30.7K engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing