A quiet supporter of general insurrection. Teacher at @FCTNOVA. Philosophy and/of Social Sciences, very broadly. Comics, Sci-Fi, Printed Books and Collapse.
New preprint: do commercial LLMs endorse ethnonationalist pseudo-science?
Grok's default (Fast) versions scored it 70-75, two to five times higher than Claude, GPT, and Gemini.
https://t.co/nQLRlfse7s
Quer queiramos quer não, a IA transforma os nossos dias. E o tempo para a debater é agora.
Nos dias 26 e 27 de setembro, todos os caminhos vão dar ao Inferência, o primeiro grande evento do Shifter. 🤖
🔗 https://t.co/tzOAJ6Aiio
Oh, btw, we sued Facebook, Instagram, TikTok & YouTube.
Reuters: Portuguese group D3 sues Meta, TikTok and YouTube over addictive design
https://t.co/blcYYlhjoh
Oh, btw, we sued Facebook, Instagram, TikTok & YouTube.
Reuters: Portuguese group D3 sues Meta, TikTok and YouTube over addictive design
https://t.co/blcYYlhjoh
@GaryMarcus This must be the only industry in which if you're building something that "does not behave the way it was supposed to" then you're doing a great job.
BREAKING: OpenAI might have stolen another major proof.
In a detailed Mastodon post, which I report in full in the comments, Andreas Thom presents several pieces of evidence suggesting that OpenAI may have trained Astra on conversations in which he and Gábor Kun were working on Gromov’s soficity conjecture, one of the ten problems OpenAI later announced Astra had solved.
I know Andreas. We met several times early in our careers. He is an exceptional mathematician, a leading expert on sofic and hyperlinear groups, and one of the most respected scholars in the field. He has spent two decades working on this problem.
If his account is correct, this is not a minor dispute over attribution. It would mean that unpublished human work was absorbed into a model and then presented to the world as a breakthrough by the model itself.
And if the allegations raised by Levent Alpöge, Tristan Buckmaster, and now Andreas Thom are all substantiated, we are no longer looking at isolated incidents.
We may be looking at one of the greatest intellectual scandals in the history of science.
AI is not discovering new mathematics.
AI is stealing human discovery.
@Bruno__Dias@LGalrao@Now_Canal_@ERC_PT Então é, ou "em Ceuta está tudo bem", ou podemos mentir descaradamente e exacerbar o medo para puxar uma agenda política que não vai muito além da xenofóbia?
O 'analista' Ireneu Teixeira assegura que são imagens 'das últimas horas em Ceuta' e que o que lhe dizem 'é que são migrantes'. NÃO SÃO. É um vídeo de Toulouse, França, após a vitória do PSG. Como é que estas vigarices são possíveis num canal de notícias, @Now_Canal_ e @ERC_PT?
Investigadores pegaram em cenários de fóruns e compararam as respostas de chatbots com as de humanos. A pergunta de partida era simples: “quão bajuladores são estes modelos?”.
E a resposta foi clara: “bastante”. 🫠
Lê mais sobre este fenómeno aqui: https://t.co/3r4gCRZphj
@fmr_porto@fcancio@poiares83 Mudanças de metodologia: em 2015 o RASI integra oito forças policiais em vez de três; em 2024 a extorsão passa a ser contada desagregada, com a extorsão sexual à parte. Nenhuma família de crime entrou ou saiu do catálogo.