Stefanie C. da Silveira
banner
ssilveira.bsky.social
Stefanie C. da Silveira
@ssilveira.bsky.social
160 followers 200 following 69 posts
Professor at Journalism Department of UFSC/Brazil. PhD in Communication Science. Digital journalism researcher. https://www.researchgate.net/profile/Stefanie-Silveira Keep on rocking in a free world.
Posts Media Videos Starter Packs
Outro caso: precisava descobrir o autor de uma citação. Joguei entre aspas no Google e ele foi incapaz de me entregar o link do artigo original mesmo estando numa revista on-line. Ele insistia em debater o texto da citação na versão IA de resultado. Acabou o buscador!
Reposted by Stefanie C. da Silveira
🚨 Virada em curso 🚨
💡Artigos escritos por IA superam textos humanos. Em 2025, 53,5% dos textos na web são IA. Áudio/vídeo seguem atrás.

💡Muda também o modelo das empresas de IA. Antes IA era ferramenta prática. Agora empresas querem "engajamento".

🕯️Análise: www1.folha.uol.com.br/colunas/rona...
www1.folha.uol.com.br
Essa semana fiz uma pergunta ao Deepseek. Qdo ele respondeu, pedi as referências da resposta. Ele deu. Tudo inventado, claro. Pedi referências reais e verificáveis, ele pediu desculpa e ofereceu alternativas. Adivinha: ele inventou de novo e ainda CRIOU LINKS FALSOS para eu “conferir” q eram reais 😫
"O chatbot do Google, aliás, foi o que se saiu pior no teste, com 76% das respostas apresentando algum problema significativo, especialmente no uso de fontes. O dado é mais que o dobro dos outros assistentes. Em seguida no ranking, vêm Copilot (37%), ChatGPT (36%) e Perplexity (30%)."
Chatbots têm erros graves como fonte de notícias 45% das vezes, diz estudo
Ao todo, 81% das respostas dos robôs apresentaram pelo menos um problema
valor.globo.com
Reposted by Stefanie C. da Silveira
The richest man on earth owns X.

The second richest man on earth is about to acquire TikTok and his family could soon own both Paramount and Warner Bros.

The third richest man owns Facebook, Instagram, and WhatsApp.

The fourth richest man owns The Washington Post.

See the problem here?
Apenas façam isso: cliquem no link. Acessem a timeline de major events e sigam rolando até o período mais recente. É de gargalhar alto!
Reposted by Stefanie C. da Silveira
Delighted to learn that Adobe, Nikon, and Leica have launched this effort to create an encrypted signature that distinguishes real digital photographs from AI imagery, and preserves the chain of custody, and that Nikon and Leica now have cameras with it.
Content Authenticity Initiative
Join the movement for content authenticity and provenance. The CAI is a global community promoting adoption of the C2PA Content Credentials standard.
contentauthenticity.org
Ponto pra ciência. Foda-se a Netflix.
O tributo é a principal fonte do Fundo Nacional de Desenvolvimento Científico e Tecnológico (FNDCT), que financia políticas de ciência e inovação no Brasil.

📰 Infos de CNN Brasil
No ritmo dos avisos de Breaking News que chegam do NYT, acredito que até o fim do ano o trump vai ter destruído tudo que resta dos EUA. Lembro de viver uma situação parecida qdo estávamos sob bolsoluço. Bóra garantir que isso não se repita mais por aqui.
Reposted by Stefanie C. da Silveira
Reposted by Stefanie C. da Silveira
Geradores de lero-lero servem para gerar lero-lero. Cura do câncer é algo bem mais sério e mais complicado do que isso. Eles sabem que a tecnologia deles para isso não serve. Mas continuam vendendo que sim.
OpenAI Wants to Cure Cancer. So Why Did It Make a Web Browser?
The AI giant has lost its imagination.
www.theatlantic.com
Reposted by Stefanie C. da Silveira
📜 CONGRESSO LOBISTA

💼 Uma investigação do UOL revelou que pelo menos 2.000 propostas legislativas apresentadas no Congresso desde 2019 foram escritas por lobistas. Há projetos de lei, emendas e requerimentos assinados por deputados e senadores, mas redigidos por representantes de empresas.
Reposted by Stefanie C. da Silveira
🚨SCOOP🚨 DHS ordered OpenAI to identify a user who'd made 2 specific prompts.

User was anonymous and accused of helping run 15 child abuse sites. They'd disclosed the prompts to an undercover agent.

First example of what I'm calling a reverse AI prompt search.

www.forbes.com/sites/thomas...
OpenAI Ordered To Unmask ChatGPT User Behind 2 Prompts
First known warrant of its kind reveals the government can ask OpenAI to provide identifying information on anyone who enters specific prompts.
www.forbes.com
Voltou o discman
CDs are enjoying a resurgence, and so are portable CD players. nyti.ms/4h7h3w6
Reposted by Stefanie C. da Silveira
You can always count on legacy media to ventilate 'both sides' of fascism.
To ensure there is 'balance' between truth and propaganda.
To pretend that they are dispassionate observers,
rather than enablers of the demise of democracy.
Reposted by Stefanie C. da Silveira
Eu sabia q a bolha da IA q estava sustentando artificialmente a economia nesse governo Trump, mas nesse nível?????? 😱
And this:
92% of the growth in the entire country this year has been data centers. And some of that is "circular deals" that dont involve actual money

fortune.com/2025/10/07/d...
Without data centers, GDP growth was 0.1% in the first half of 2025, Harvard economist says | Fortune
Is a U.S. without data centers a country without GDP growth?
fortune.com
Reposted by Stefanie C. da Silveira
Reposted by Stefanie C. da Silveira
Studies are beginning to reveal fissures in the argument that AI boosts productivity, writes Tech Policy Press fellow @eryk.bsky.social. Nonetheless, companies and governments are pouring investments into speculative growth without ample evidence.
Generative AI’s Productivity Myth | TechPolicy.Press
People may be using artificial intelligence, but that doesn’t mean it’s useful, writes Eryk Salvaggio.
www.techpolicy.press
Reposted by Stefanie C. da Silveira
“Current architectures of LLMs cannot imagine, but they can sequence … For the same reason that a dog can go to church but a dog cannot be Catholic, an LLM can have a conversation but cannot participate in the conversation.”
I've been re-acquainting myself with NNs, transformers and LLMs this week, so these are early thoughts. But I am starting to think about language without the capacity to imagine language, i.e., an LLM unable to imagine itself participating in language. mail.cyberneticforests.com/what-machine...
What Machines Don't Know
Imagining Language Without Imagination It's important to acknowledge that Large Language Models are complex. There's an oversimplified binary in online chatter between the dismissive characterizatio...
mail.cyberneticforests.com
Reposted by Stefanie C. da Silveira
Reposted by Stefanie C. da Silveira
1/10 There are very abstract concepts that must be taken as truth in order to subjugate humans to technology and its owners.One of the most powerful is the idea that computation is consciousness and life is information,the core belief behind what philosophers call computational functionalism.
👇
Reposted by Stefanie C. da Silveira
If you're going to write headlines like this, why even have newspapers?
Should the New York Times be sold to Pennysaver and its editors sent to work in Amazon warehouses? It depends on who you ask.
Reposted by Stefanie C. da Silveira
twice a year the entire U.S. press becomes a marketing extension of a single billionaire-owned retailer, and nobody in any position of editorial power thinks it's weird or gross