Andre

27.5K posts

Andre banner
Andre

Andre

@andre_arag

Eu rio alto - I laugh out loud

Colorado, USA Beigetreten Nisan 2009
1.7K Folgt764 Follower
Andre
Andre@andre_arag·
@kaaaash____ Left. By far. And I had 4 magic mouses before
English
0
0
0
3
Akash
Akash@kaaaash____·
Be honest, Left or Right
Akash tweet mediaAkash tweet media
English
1K
50
2.9K
650.8K
Adriana Porter Felt
I asked gemini to give my basement a glow up and then went out and bought all the actual furniture. First is the nano banana mockup; second is the actual room as it looks today; third is the original photo of the room that nano banana was working off of.
Adriana Porter Felt tweet mediaAdriana Porter Felt tweet mediaAdriana Porter Felt tweet media
English
85
43
1.6K
2.3M
Andre
Andre@andre_arag·
@leomnonato @Cardoso Pq as pessoas não podem avisar? Perguntar se pode? Isso é uma mistura de má educação com egoísmo e desprezo pelo outro.
Português
0
0
35
594
Leo Nonato 🇧🇷
Leo Nonato 🇧🇷@leomnonato·
@Cardoso Visita inesperada virou incômodo pois a tela dos dispositivos estão cheias de estranhos e é infinitamente mais interessante né verdade?
Português
26
0
163
17.1K
cardoso
cardoso@Cardoso·
Aqui vou ficar do lado do Jovem™. Isso é um hábito horrendo de antigamente que acabou graças à Democratização das telecomunicações. Eu faria o mesmo.
cardoso tweet media
Português
152
174
6.4K
120K
Andre
Andre@andre_arag·
@joaopedro @Felippe_Hermes @lylasbitch Ele teve Stock options vesting a cada 4 anos, no mínimo. Isso significa que a cada 4 anos ele deve ter ganho algo em torno de 2 a 3x o salário dela em bônus. Se ela investiu corretamente, ela não precisa se preocupar em trabalhar mais.
Português
0
0
26
843
João Pedro C. Motta
João Pedro C. Motta@joaopedro·
@Felippe_Hermes @lylasbitch kkkk pqp a mina é mto burra, não sabe o que é Oracle e nem quanto ganha alguém com esse currículo. A gente tá discutindo com uma porta. Não vale a pena.
Português
6
3
668
6.5K
Juh
Juh@lylasbitch·
A querida com 34 anos de empresa foi demitida e sabe o que é o pior? Nos Estados Unidos não existe FGTS e nem multa de 40% não existe sequer garantia de receber qualquer coisa na demissão depende muito da empresa e estado pois não é obrigatório. 34 anos e saindo de mãos abanando
Apurv Khare@KhareApurv

33 years 😮 Damn #OracleLayoffs

Português
294
55
724
453.3K
PokerDev
PokerDev@pokerdev7·
RESOLVI OS PROBLEMAS DE LIMITE DO CLAUDE CODE apenas rode no terminal: npm i -g @openai/codex
Español
9
2
194
13.8K
Sindpd SP | Sindicato de TI 💻
PESSOAL, ATENÇÃO! Recebemos denúncias de demissão em massa na ORACLE. Além da não negociação com o sindicato, a empresa apresentou um termo de “NÃO HOMOLOGAÇÃO NO SINDICATO”. A homologação é um procedimento gratuito, garantido em Convenção Coletiva para os trabalhadores. Somente com a homologação, o Sindpd-SP recuperou mais de 10 milhões de reais para os trabalhadores no último período por erros da empresas nas rescisões contratuais. Se você foi demitido ou conhece quem foi demitido na Oracle, nos acione no nosso WhatsApp ou no nosso e-mail coe@sindpd.org.br Ajude compartilhando este post!
Português
24
106
954
58K
Andre
Andre@andre_arag·
@aakashgupta A lot more stuff gets done. The demand is not slowing down. It’ll swallow all these improvements and ask for more.
English
0
0
0
47
Aakash Gupta
Aakash Gupta@aakashgupta·
The real story is the 14x compression ratio and what it means if it scales up. Every single weight in this model is one bit. Zero or one. That's it. 8.2 billion parameters stored in 1.15 GB of memory. A standard 8B model at full precision takes 16 GB. Bonsai 8B fits on your phone with room left over for your photo library. The benchmarks are the part that shouldn't be possible. On standard evals, a model that's 1/14th the size of Qwen3 8B and Llama3 8B is trading punches with both of them. The intelligence density score, capability per GB, is 1.06/GB versus Qwen3 8B at 0.10/GB. That's a 10x gap in how much thinking you get per unit of storage. Now zoom out. Big Tech collectively spent over $320 billion on data center capex last year. Amazon alone dropped $85.8 billion, up 78% year over year. Google committed $75 billion for 2025. The US power grid is buckling under AI demand. Data centers now consume 4.4% of all US electricity. Virginia, where most of them sit, saw electricity prices spike 267% over five years. Residential customers in Ohio are watching their bills climb 60% because utilities are spending billions on transmission infrastructure to feed server farms. The entire AI scaling thesis runs on one assumption: intelligence requires massive compute. PrismML just published a proof point that the assumption might be wrong. Their CEO, Babak Hassibi, is a Caltech professor who spent years on the mathematical theory of neural network compression. The founding team is four Caltech PhDs. Khosla Ventures backed it. So did Cerberus, whose Amir Salek built the TPU program at Google. The 1.7B model runs at 130 tokens per second on an iPhone 17 Pro Max at 0.24 GB. The 4B hits 132 tokens per second on M4 Pro at 0.57 GB. These aren't research demos. They shipped llama.cpp forks with custom 1-bit kernels for CUDA and Metal. Apache 2.0 license. You can download and run it right now. The trillion-dollar question: what happens to the economics of a $75 billion data center budget when the same intelligence fits in 1/14th the space and runs on 1/5th the energy?
PrismML@PrismML

Today, we are emerging from stealth and launching PrismML, an AI lab with Caltech origins that is centered on building the most concentrated form of intelligence. At PrismML, we believe that the next major leaps in AI will be driven by order-of-magnitude improvements in intelligence density, not just sheer parameter count. Our first proof point is the 1-bit Bonsai 8B, a 1-bit weight model that fits into 1.15 GBs of memory and delivers over 10x the intelligence density of its full-precision counterparts. It is 14x smaller, 8x faster, and 5x more energy efficient on edge hardware while remaining competitive with other models in its parameter-class. We are open-sourcing the model under Apache 2.0 license, along with Bonsai 4B and 1.7B models. When advanced models become small, fast, and efficient enough to run locally, the design space for AI changes immediately. We believe in a future of on-device agents, real-time robotics, offline intelligence and entirely new products that were previously impossible. We are excited to share our vision with you and keep working in the future to push the frontier of intelligence to the edge.

English
69
102
914
151K
Andre
Andre@andre_arag·
@fiBonatto Pq não tem engenharia no Brasil. So “web page”, software de padaria e “enterprise” slop.
Português
0
0
6
145
Bonatto
Bonatto@fiBonatto·
why are there no C jobs in brazil??? everything is react/ts, java, or php i want to work with C and can’t find anything in this country
English
93
10
531
31.5K
Andre
Andre@andre_arag·
@altryne @zerohedge If this rust port works I’ll be a very happy camper. Tired of so much slow node/js crap running on my machine
English
0
0
1
326
Alex Volkov
Alex Volkov@altryne·
If you, like me, just woke up, let me catch you up on the Claude Code Leak (I know nothing, all conjecture): > Someone inside Anthropic, got switched to Adaptive reasoning mode > Their Claude Code switched to Sonnet > Committed the .map file of Claude Code > Effectively leaking the ENTIRE CC Source Code > @realsigridjin was tired after running 2 south korean hackathons in SF, saw the leak > Rules in Korea are different, he cloned the repo, went to sleep > Wakes up to 25K stars, and his GF begging him to take it down (she's a copyright lawyer) > Their team decided - how about we have agents rewrite this in Python!? Surely... this is more legal > Rewrite in Py > Board a plane to SK🇰🇷 > One of the guys decides python is slow, is now rewriting ALL OF CLAUDE CODE into Rust. > Anthropic cannot take down, cannot sue > Is this "fair use?" > TL;DR - we're about to have open source Claude Code in Rust
Alex Volkov tweet media
English
357
1.1K
11.9K
2M
Nick
Nick@NicAnkush·
@kenzietuff But why is the media obsessed with this topic?
English
10
0
13
8.3K
Mack
Mack@kenzietuff·
If you’re a wife you need to understand 99% of media you see is trying to convince you to hate your husband/children. If you cannot handle this and stay grounded, log off. Constant stream of anti-family, spousal resentment propaganda. Most modern divorces are a result of this.
English
312
5K
30K
611.2K
Andre
Andre@andre_arag·
@DailyMail LOL - u guys are such losers
English
0
0
0
5
Daily Mail
Daily Mail@DailyMail·
Having children DOESN'T make you happy: Becoming a parent provides no boost to wellbeing, study claims - but it does ruin your relationship trib.al/VHoMJzK
English
938
169
733
458.3K
None None
None None@Nebraskaman9694·
@TheBabylonBee And being incredibly socially awkward among his age group.
English
11
0
6
435
Andre
Andre@andre_arag·
@barraes Eles estão subsidiando o custo em 25x1. Vai piorar.
Português
1
0
3
1.1K
Bruno
Bruno@barraes·
Então rapaziada claudera do X... Eu tenho tido esse problema A SEMANA TODA. Não tá legal não, e eu faço prompts em ingles, enxutos (até demais as vezes) só que não to conseguindo trabalhar 3 horas seguidas, por exemplo... Estão nessa? Estou pulando entre Codex e Gemini o tempo todo! O Claude tá comendo tokens como se não houvesse amanhã, e ainda diminuiram os limites...
Sai Satvik@saisatvik_

Just 2 prompts. The input was around 300 characters in total, and the output was around 700-800 characters in total. It feels like @claudeai is just scamming with credits. Thought to upgrade to the $200 plan, but now it feels like it's not worth it.

Português
52
3
208
42.6K
Andre
Andre@andre_arag·
@giovannibassi Não posso dar detalhes mas é um complex react frontend com um backend em golang também razoavelmente complexo.
Português
1
0
1
43
Andre
Andre@andre_arag·
@fabiomoraes Sem contar q ele reclama se vc não descolore
Português
0
0
1
84
Andre
Andre@andre_arag·
@pedreco Treno. Trenar. Esse pessoal não usar o i
Português
0
0
1
59
Pedro Aguiar
Pedro Aguiar@pedreco·
Quando foi que começaram a falar "treinar" em vez de malhar? Pq pra mim treino é algo que se faz pra se preparar pra um esporte ou uma competição. É coisa de atleta. Quem faz exercício por saúde ou estética tá malhando, não treinando.
Português
262
126
1.8K
78.3K
Andre
Andre@andre_arag·
@brunooomelo Continuo usando mesmo com Claude code
Português
1
0
0
71
Andre
Andre@andre_arag·
@MsAmericanPie_ Absurdo - sinto muito. Minha família tem muitos professores (incluindo meu pai, irmão e irmã) e aprendi desde pequeno o q significa isso.
Português
0
0
1
14
Taina
Taina@MsAmericanPie_·
@andre_arag Mas falam, e falam muito
Português
1
0
0
43
Taina
Taina@MsAmericanPie_·
Lecionar 10 períodos de aula das 7 às 18h sem pausa pra ouvir “tu só dá aula ou trabalha também?”
Português
5
1
35
907