JacobFinite

Jak dostaneme chudé z chudoby? 1. Stát by měl přestat brát ~70% výplaty skrz daně. 2. Stát by měl omezit počet regulací, díky kterým bude volný trh fungovat efektivněji a pomůže vytvořit obecně více bohatství a lepší životní úroveň pro VŠECHNY.

@wokeancap @TadeasuvSpace Jsou to peníze který pak nemůžou přijít na státní terorismus, jako jsou třeba ICE nebo obdobný aktivity; za mě je to lepší alternativa.

@TadeasuvSpace Ano, je to dement, když navrhuje a souhlasí s tím, že se budou dávat prachy na výstavbu naprosté zbytečnosti pro arogantní elity.

Pěknej čurák se z něj stal.

@VictorTaelin @blingdivinity exactly, it seemed so dumb to me but if it works it works lol

@blingdivinity even worse, habits learned from a model don't work in the other. like I gave it a full spoonfeed prompt thinking that would help but turns out "replicate this in an angle pls" is what makes it click

kinda funny that despite claims that we’ve graduated to “context engineering”, prompt engineering is still a thing 3 years after chatgpt

@Patriksailor @TadeasuvSpace Není to že je heterák o tom s kým chce spát on, a ne o tom s kým jeho partnerka?:D

@TadeasuvSpace Heterák tohle rozlišuje.

Přijde vám normální, aby chlap dělal scény kvůli tomu, že ho žena “podvedla” s jinou ženou? Ptám se vážně, protože asi nejsem ten nejpovolanější k tomu, abych to dokázal přiteli nějak rozumně vysvětlit.

@subzndubz @mayfer I expect you don't get IV injections in the hospital, even for a billion dollars, right?

@mayfer If having a billion dollars means I have to inject chemicals I’ll pass. Now put that into a smokable or edible and I’ll be a billionaire right there with you.

if you find a cool new psychedelic with ideal trip duration you can make a billion dollars apparently

Tohle mi příjde jako tak strašně nonsensical hoax; kdokoliv, kdo chápe jak funguje realita (voda magicky nemizí po použití) a počítače (počítače nepijou ani neznečišťujou vodu), to vidí na první dobrou. Většina mých good faith leftist kamarádů tomu ale seriozně věří.

@oldasadlo @VratislavSvejda @frvavra Státní fízl, kterej by mě s naprostým klidem násilím odvlekl za mříže za věci, kterýma nikomu neublížím (jako třeba v případě Kordysových); to jsem "já" nebo "my" nebo "nasi fizli"? Kde si přišel na to že stát jsem "i já"?

@JacobFinite @VratislavSvejda @frvavra stat jsme my, to jsou nase krochny a nasi fizli. neni to moc dobry, ale kdyz se budem soustredit a snazit se, muzem si mozna udelat lepsi stat. bez statu ale nemuze existovat zadne pravo a bez prava nemuze byt jakakoliv spravedlnost nebo rovnost, jenom bezpravi a dominance.

Krize bydlení je zcela umělý problém způsobený státem. Kdyby si každý na vlastním pozemku mohl prostě postavit, co chce*, bylo by po problému. Jenže na pozemek i dům potřebujete od státu PAPÍR, který to mnohonásobně prodražuje. *a neškodí tím ostatním x.com/i/status/20339…

@oldasadlo @VratislavSvejda @frvavra Ale vždyť ve světě kde "silnější pes mrdá" furt žijeme; stát je ten silnější pes tady, má armádu, policii, a spoustu zbraní. Co je na tom spravedlivýho? Že můžeš hodit papírek do krabičky jednou za čtyři roky?

@VratislavSvejda @frvavra dava to smysl, libertariani chteji z principu mene spravedlivy a hierarchizovanejsi svet kde silnejsi pes mrda. asi si mysli ze jsou bohati, jinak mi to nedava smysl.

@VictorTaelin I quite enjoy GLM 4.7 flash for local, but honestly the small models are just too dumb to be worth the hassle :D

So what's currently the best OSS coding model that can run on the base Mac Mini, to be used with Pi? Is there anything decent? Is it Qwen3.5-9B?

@maxviolense @HonzaPoboril @VojtaZizka Most likely not; spíš si myslím že runnujou jen openclaw s LLMs od providers :D

@HonzaPoboril @VojtaZizka Ak teda plati že na tom runuje local LLMS

Tímhle tempem budu mít za chvíli v týmu víc botů, než lidí.

@VictorTaelin Epic progress!!

Claude and Codex are now competing to prove the correctness of a small vibe-coded app in Bend2 🥳

@VictorTaelin @itaybachman Sending love, you'll figure this shit out like you always do <3

@itaybachman I'm not in a good mental state to care about that ):

just thinking out loud about all the hard things I still need to do before launching Bend2, and consequences if we launch without each 1. HVM4's AOT compiler that means compiling the HVM4 efficient C. that gives a 10x-100x speedup in practice, so, it is essential for the HVM4 back-end to be viable. without it, you'd be using 10's of threads to run an interpreter. that was the main criticism of Bend1. given that parallelism was a key feature of v1, I think we *need* a really good version of it, even though it isn't the central point anymore. we have an initial compiler, but would probably take 5-10 days of focused work to make it good 2. HVM4's GPU runtime it still doesn't have one, at all. doing so is way more complex than HVM2, because of lazy (rather than super strict) evaluation. launching without a GPU runtime would feel like a regression and this is basically locking me. probably 15 days of work 3. Bend2-SupGen integration this is really hard to do, due to Bend's dependent types. currently, SupGen is still reliable for simple types. without it, the time/cost it takes for the AI to fill all sub-proofs is still too high for it to be a viable product. we're talking about whole nights to make a small refactor. probably 5-10 days of focused work ... and that's basically it actually. if these things were ready, all the rest is trivial, and we could launch it the next week... I wonder if I should just launch Bend2 without the HVM4 back-end. keep in mind that Bend2 compiles to very fast JS, so we don't *need* HVM4 or parallelism in any way. but it would be extremely hard to communicate that, given the initial appeal of the language was parallelism and interaction nets. we have small windows where people take the time to try our stuff, so I think if it is not great, polished and covers all corners on launch, it will just be dismissed as a "not ready yet" thing for another year not sure the "move fast and break things" advice applies to when you're building a programming language, Bend1 did launch prematurely and I shouldn't let that happen again. perhaps "a delayed game is eventually good, but a rushed game is forever bad" is the advice I should be following here...

@VictorTaelin Genuinely just use OpenCode. You can roll your own plugins, system prompts, everything; and you can also very easily interact with it using the Typescript SDK, if you want cool custom stuff. The TUI is gorgeous too btw ;)

I just want a Claude Code that is much more lightweight. Way smaller system prompts, way less bloat in general. It could probably be 3x faster and more effective if it wasn't overly bloated. This is really frustrating. Perhaps I should roll my own

@norpadon What font is that? So gorgeous

Codex was here

@VictorTaelin @buildwithbc The Claude plugin sucks btw! Doesn't really do what Ralph is supposed to. I recommend looking at either AI hero explaining it, or it's creator explaining it, here: aihero.dev/getting-starte… ghuntley.com/ralph/ Been using it on my projects and it's great.

@buildwithbc I see! I missed that wave then

Just woke up and checked it! Overnight: → 3 major bugs fixed → all benchmarks improved (from 4.8% to 92.8% !!) → total code size *reduced* by 10% I just set up a GOAL.txt with optimizable metrics and left a crowd of AI's improving the Ben2 repo. Very cool

@dkubb @VictorTaelin I'm excited for the GPT 5.3 spark to be available on API, as that will make these subagents so so fast.

I'd suggest to have Opus delegate some of the work to a Haiku subagent and have it watch and measure how the subagent does. I often have Opus create tests that lower-powered agents have to carry out in order to measure how easy to use something is, or how well documented it is. I'll iterate on the docs or UX until the tool can be used effectively by the weaker model.

right now Codex and Claude are writing dozens of Bend2 benchmarks (sort, nqueens, huffman, etc.) as I go to bed I also asked them to keep a `feedback.txt`, where they report on their experience with Bend2: bugs, unhelpful errors, missing features, etc. Claude already has a bunch of complaints, it seems no worries tho, it will be tasked with fixing them all tomorrow 🥰 good night!

@FebrezeSamurai @Liberty_Tyrant Does Zulu not argue based on economics?

@JacobFinite @Liberty_Tyrant He’s based but his argumentation is only useful if you’re debating against philosophy autists. PraxBen argues based on economics and Mentis is kind of a generalist (he puts some care into optics as well)

*sigh* PraxBen seems like a whiny bitch. MentisWave keeps pandering to rightards. I consider Taxed Brass as kind of a homie on Twitter, but in truth, I haven’t really watched any of his videos. In truth, I think there is only like 6 actual Hoppeans. Hoppe is no dumbie, so his epistemology and ethics are quite dependent on each other. Not many people seem to understand this.

@VxalCZ Proč?

@JacobFinite Mám v plánu vyzkoušet, ale z toho co jsem zkoušel CC jednoznačně vede.

Můj pracovní den Claude Code s jádrem Opus 4.6 dělá práci, kterou následně kontroluje CC s jádrem glm-5 a kimi-k2.5 mající za úkol brblat a dokázat hlavnímu agentu že je dement. Jsme dobrá pracovní skupina.

@MichaelHauda @sebascespedes98 Because you can't alienate yourself from the property you are trying to sell (your body), which is required for a title transfer of property.

@sebascespedes98 Why shouldn't I be allowed to if I want to sell myself into slavery? What right do you have to say no?

Slavery is when you voluntarily sign a contract to work more hours than a communist thinks you should be allowed to (you've gone against party policy and will now be worked to death in Siberia)