Reposted by Antonio Pedro ☕Inimigo do Rei @cyrusafa.bsky.social · 09/10/2026BORA ESPALHAR ENTÃO NA TIMELINE DE CADA UM DE NÓS Se Flávio ganha, você perde 27669455
Antonio Pedro ☕ @apsv.bsky.social · 07/10/2026as vezes a pessoa tem que ter um pouquinho de determinação e lembrar que não é invertebrada 000
Reposted by Antonio Pedro ☕Celso Rocha de Barros @npto.bsky.social · 04/10/2026Bora lá galera, boa sorte pra todo mundo que não for um fascista arrombado do caralho. 212085419
Antonio Pedro ☕ @apsv.bsky.social · 02/10/2026parabéns para a tv Globo que mesmo sem o debate conseguiu favorecer os seus 000
Antonio Pedro ☕ @apsv.bsky.social · 30/09/2026the amount of software you can get for 5 bucks is just crazy 000
Reposted by Antonio Pedro ☕diabo destruidor de mundos @pedrofonseca.bsky.social · 29/09/2026 322241957
Antonio Pedro ☕ @apsv.bsky.social · 29/09/2026let's please not adopt the "System One" nomenclature 001
Reposted by Antonio Pedro ☕Jeff Nascimento @jnascim.info · 23/09/2026Meta derruba perfil de série com Porchat sobre relações entre Master e Igreja Lagoinha. Teaser somava 6 milhões de visualizações antes da retirada da conta jnascim.info/mM9HpPjnascim.infoMônica Bergamo: Meta derruba perfil de série com Porchat sobre relações entre Master e Igreja LagoinhaPlataforma alega violação de padrões ao canal Revelando, mas não detalha regra descumprida; ela não responde à coluna 55914300
Antonio Pedro ☕ @apsv.bsky.social · 21/09/2026i can do this in bash, thus there's no need for abstraction 000
Reposted by Antonio Pedro ☕Viada, puta e comunista @priskika.bsky.social · 21/09/2026Infelizmente o PCC arrombou o escritório do Milton Leite e levou tudo que poderia comprovar a ligação do ex-presidente da câmara de vereadores ao crime organizado. Uma pena que a polícia não tenha conseguido cumprir o mandado de busca e apreensão que saiu na semana passada. Coisas da vida. 25899230
Antonio Pedro ☕ @apsv.bsky.social · 19/09/2026aren't most of these totalitarian exploitation scenarios where people work in inhumane conditions, only the degrading jobs that aren't automated yet? idk sounds familiar 1140
Reposted by Antonio Pedro ☕Spooky Key 🎃 🗝🦊✅ @keytryer.net · 16/09/2026That's right, it goes in the JSON hole 117719
Antonio Pedro ☕ @apsv.bsky.social · 16/09/2026the weirds are pretty loud, makes the whole thing tiresome but anyway, what else can we do 010
Antonio Pedro ☕ @apsv.bsky.social · 16/09/2026let me just start by saying minds don't actually have two systems okay 000
Antonio Pedro ☕ @apsv.bsky.social · 12/09/2026I was wrong, in the start of 2025, when I thought LLMs would stagnate and any gains would come only from more compute being thrown at each task. Thinking that now is pretty much insane... And even then, if LLMs only improve with more compute, you'd have to be crazy to bet that they won't do it. 000
Antonio Pedro ☕ @apsv.bsky.social · 12/09/2026turns out we're all gonna be hacked because of a location sharing app nobody liked 000
Antonio Pedro ☕ @apsv.bsky.social · 11/09/2026it's only a "sandbox escape" if the computer phisically gets out of the sandbox and goes to play on the street 0142
Antonio Pedro ☕ @apsv.bsky.social · 10/09/2026I'm so glad Anthropic has thought about the economy and found out only good things can happen www.anthropic.com/institute/ec...anthropic.comScenarios for our Economic FutureThe Anthropic Economics Team models the effects of AI on the economy of 2030. 000
Antonio Pedro ☕ @apsv.bsky.social · 09/09/2026look, if and when we get through alignment... how the fuck will we get through curtailment for misaligned groups training models aligned to their intent 000
Reposted by Antonio Pedro ☕dame @dame.is · 09/09/2026we need woke gay solar powered data centers, and we need them now 715218
Antonio Pedro ☕ @apsv.bsky.social · 09/09/2026There is a version of the future where we look back on this period like we do on the Atom Fever times. On the other versions I assume the comparison is a bit too dark 000
Antonio Pedro ☕ @apsv.bsky.social · 09/09/2026Assumes reality is Wiener Process and sacrifices one reality for the good of many 020
Reposted by Antonio Pedro ☕Olúfẹ́mi O. Táíwò @olufemiotaiwo.bsky.social · 09/09/2026applying "Make ___ Great Again" to Germany really does kinda give the game away 723886720
Antonio Pedro ☕ @apsv.bsky.social · 09/09/2026i heard anthropic is about to publish a breakthrough in package management 000
Antonio Pedro ☕ @apsv.bsky.social · 08/09/2026hey folks is now a good time to talk about funding for ai infra in universities and research institutions 000
Antonio Pedro ☕ @apsv.bsky.social · 08/09/2026toda vez que eu escuto "andré do prado" eu penso no "andré do rap" 000
Reposted by Antonio Pedro ☕konnaire @konnaire.bsky.social · 08/09/2026Putting the :/ in https:// 92473785
Antonio Pedro ☕ @apsv.bsky.social · 08/09/2026if we do more, but understand less of what we do, are we not all executives whose "vision" and perceived guidance is evaluated on expected results and politics? 000
Reposted by Antonio Pedro ☕Simon Willison @simonwillison.net · 07/09/2026"I don't know the answer myself, but I asked a blowhard I know and he took a wild guess, here's what he said: " How I interpret pasted replies from an LLM in online conversations 1616620
Antonio Pedro ☕ @apsv.bsky.social · 06/09/2026just needs to be a group large enough to pay for b200s 000
Antonio Pedro ☕ @apsv.bsky.social · 06/09/2026community hosting of AI applications sounds pretty good: more technical people set up self-hosted models and apps, pricing can essentially shared cost, privacy and trust are built between the people... 120
Antonio Pedro ☕ @apsv.bsky.social · 05/09/2026honestly, LLM posts I can take, but LLM messages from people you can't ignore are just too far 000