We’re rolling out Gemini 4 Argon in Antigravity to paid Google AI Pro subscribers today. Demand has been incredible and we've scaled capacity deliberately to get here. We’ve rolled out a separate weekly-only Argon quota for Pro on top of the existing Gemini bucket for 3.8 Flash so you have the best of Gemini for all your tasks.
1/9 E se eu falar que eu treinei um LLM do Zero, com dados 100% br, e feito sem gastar 1 real?
Hoje decidi abrir publicamente o código, os dados e os pesos do meu projeto de LLM Legislativo Brasileiro. É um pipeline completo que vai desde a coleta de quase 1 milhão de documentos históricos do Congresso Nacional até o pré-treinamento de um Transformer causal próprio.
Não usei pesos pré-treinados de ninguém. Não usei APIs fechadas. Não usei atalhos prontos como o Hugging Face Trainer. Fiz tudo do zero em PyTorch puro, com foco em reprodutibilidade e integridade matemática.
Vou contar como estruturei isso no fio abaixo.
Se a gente for ver os dados de custo de coalizão, como medidos por Carlos Pereira e Frederico Bertholini, esse foi o governo com maior sobrerrepresentação do partido do presidente na história da Nova República.
Vocês podem não gostar da realidade, mas ela é essa. E sim, houve uma depletação da Frente Ampla já na transição, que se traduziu não só na política macroeconômica, mas na externa, no conflito com o Banco Central, na Política Industrial... tudo isso, tirando a política externa, afetou diretamente o resultado eleitoral, e isso também é um dado concreto da realidade.
E mais, não só havia uma sobrerrepresentação do PT como os próprios ministros e burocratas do PT não tinham acesso ao presidente. Parte significativa dos problemas com a Taxa das Blusinhas e a explosão cambial e dos juros de 2024 poderia ter sido evitada se Lula escutasse Haddad sem ter que passar pelo filtro de Rui Costa, Janja e Sidônio que já crepitava no planalto.
Precisamos radicalizar a esquerda para o próximo ciclo político!
Chega de Tebet, João Campos, Haddad, Eduardo Paes e outros "moderados" ou que se aproximam somente por oportunismo eleitoral!
+ Jones Manoel
- Tábata Amaral
Eu não acredito mais nesse hype que o Tibo faz. OpenAi se fudeu, hypou o DevDay e não entregou nada. Se acham melhor que a Anthropic (por terem mais usuários) e só entregam produto meia boca. Nem dumping conseguem fazer direito, boa sorte no IPO
Massive L for OpenAI today.
Full of disappointing announcements, presentation problems, and nothing exciting.
Plus a $500 plan.
Claude is eating them alive right now.
Hi,
Tomorrow we are re-opening the Pro $200 subscriptions to new subscribers, but together with it we are also changing how we calculate the usage for it. In effect, if you do the math, it will net out at half the dollar in API spend compared to the old Pro $200 plan.
Now that it's said, let me explain why this is happening and why you will still get more work done than if you were on the Pro $200 subscription one month ago.
(a) We didn't want to compromise in other ways and are committing to not reintroducing the 5h limit, so that you can fully use the weekly usage when you want.
(b) On the subscription, we guarantee that over time you always get more work done and with an increasing level of quality. This means that you will continue to get more value per dollar spent as a result of models getting more efficient and us passing down the improvements in the form of API price reductions.
(c) We don't want to put an incentive on ourselves to artificially inflate the API list prices to make it look like you are getting a lot (and workaround it through discounts, etc). Instead we want to continue to both rapidly reduce prices and increase capabilities of models on the API. This week we introduced GPT-6 Sol and GPT-6 Luna at 50% of their previous price. Over time, we see prices go low enough that it makes sense for most to buy usage as needed without there being a significant gap between what you get in a subscription and what you get in the API for a dollar spent.
(d) Tomorrow, we are adding more things to the subscription that won't draw on the usage, I won't reveal what that is yet.
I wanted to be transparent before all the big announcements tomorrow. Lots of new exciting things are coming to the subscriptions that will make it super compelling, but I wanted to make sure to share this change ahead of time so you can all understand it before we shower you with good news.
Codexingly,
Tibo
In the Democratic Republic of the Congo, global health organizations including @CEPIvaccines, @WHOAFRO, and @inrb_kinshasa are using Claude to accelerate their response to an outbreak of an unusual Ebola variant.
Read the full piece here: https://t.co/lQKx17Mxc3
🇧🇷 AGORA: Ministro Alexandre de Moraes se declara impedido de julgar mandado de segurança que trata sobre sua própria esposa, Vivivane Barci de Moraes, e o ex-banqueiro Daniel Vorcaro.
If I had to describe Brazil’s election in 2 words, they’d be “apathetic polarization.” Yes, it’s (still) a 50-50 country, but the clear vibe of the 2026 election is the lack of enthusiasm on either side. You feel it the moment you arrive — so hard to find anybody who’s truly EXCITED to vote for Lula or Flávio. So different from 2018 or 2022 when Brazilians felt there was so much at stake — because of the bad economy, the pandemic and the threats to democracy. This time around, voters say meh, the economy is OK but my salary seems worth less, Lula is OK but old, Flávio has skeletons but is not perceived as aggressive as his dad.
In this… windless environment even the slightest breeze can seem relevant. Flávio is currently having a moment in polls, it’s probably because of what’s happening in Supreme Court, but I’ve spoken to analysts who wonder if it’s actually voters reacting to higher fuel prices from Iran war and overall inflationary pressures and household debt. It’s a VERY hard election to call because it’s simultaneously very close but the level of conviction among many voters is so low. Flávio slight favorite today but for now I’d avoiding making big bets either way.
PS “apathetic polarization” is a bit academic for me but the best I could do 👍🙏