Bruno Amaral
@brunoamaral.eu
I want the web to be a better place. Communication Strategist, Creative Tinkerer.
Weirdness and occasional posts in Portuguese.
https://brunoamaral.eu/
Helping Multiple Sclerosis research at @gregory-ms.com
Founder of the Lisbon Collective consulting.
Weirdness and occasional posts in Portuguese.
https://brunoamaral.eu/
Helping Multiple Sclerosis research at @gregory-ms.com
Founder of the Lisbon Collective consulting.
Reposted by Bruno Amaral
When people saw more divisive posts on X they spent more time there but liked and shared less, and felt colder towards their political opponents. Fewer divisive comments and they spent less time on X but liked and shared most posts. So platforms COULD CHOOSE TO reduce/stop increasing polarisation.
November 28, 2025 at 10:53 PM
When people saw more divisive posts on X they spent more time there but liked and shared less, and felt colder towards their political opponents. Fewer divisive comments and they spent less time on X but liked and shared most posts. So platforms COULD CHOOSE TO reduce/stop increasing polarisation.
the emoji for receipt is funny.🧾
November 28, 2025 at 4:24 PM
the emoji for receipt is funny.🧾
Malta, já sei! Em vez de investir em mais alojamento local, transformamos os apartamentos de Lisboa em pequenos data centers. No verão é chato, mas no outono e inverno ninguém vai ter problemas de aquecimento e o aluguer do espaço paga a conta da luz!
November 26, 2025 at 5:47 PM
Malta, já sei! Em vez de investir em mais alojamento local, transformamos os apartamentos de Lisboa em pequenos data centers. No verão é chato, mas no outono e inverno ninguém vai ter problemas de aquecimento e o aluguer do espaço paga a conta da luz!
What ChatGPT says about me “You blend engineer, designer, writer, sysadmin, professor, and art director into one continuous stream — and expect the AI to keep up without blinking.”
November 25, 2025 at 10:34 PM
What ChatGPT says about me “You blend engineer, designer, writer, sysadmin, professor, and art director into one continuous stream — and expect the AI to keep up without blinking.”
Everybody loves innovation until they see pineapple on pizza.
November 24, 2025 at 6:54 PM
Everybody loves innovation until they see pineapple on pizza.
As pessoas que nunca apoiaram os artesãos e artistas locais agora estão muito indignadas com a inteligência artificial.
November 23, 2025 at 3:38 PM
As pessoas que nunca apoiaram os artesãos e artistas locais agora estão muito indignadas com a inteligência artificial.
Screego
screen sharing for developers
app.screego.net
November 20, 2025 at 6:36 PM
Reposted by Bruno Amaral
The Digital Omnibus proposes to deregulate privacy & AI in Europe. It is a terrible idea. I'm one of the signatories of an open letter to the Norwegian government explaining why we need to keep GDPR, the AI Act and other tech regulation in place. Especially now. www.forbrukerradet.no/siste-nytt/d...
Forslag om "forenkling" av digitale regelverk svekker forbrukerrettigheter
Nytt lovforslag er en avvikling av grunnleggende rettigheter, mener Forbrukerrådet og en rekke andre organisasjoner og privatpersoner.
www.forbrukerradet.no
November 18, 2025 at 12:39 PM
The Digital Omnibus proposes to deregulate privacy & AI in Europe. It is a terrible idea. I'm one of the signatories of an open letter to the Norwegian government explaining why we need to keep GDPR, the AI Act and other tech regulation in place. Especially now. www.forbrukerradet.no/siste-nytt/d...
Brands in 2008: We want to be on social media and be more authentic!
Brands in 2025: We want to use AI to make better ads for social media!
Brands in 2025: We want to use AI to make better ads for social media!
November 18, 2025 at 11:40 AM
Brands in 2008: We want to be on social media and be more authentic!
Brands in 2025: We want to use AI to make better ads for social media!
Brands in 2025: We want to use AI to make better ads for social media!
Double-edged sword, @gregory-ms.com got new training data and became more accurate in finding papers with hope for people with #MultipleSclerosis. Visits to the website are down because those papers don't show up as often. #MSchat
November 18, 2025 at 10:09 AM
Double-edged sword, @gregory-ms.com got new training data and became more accurate in finding papers with hope for people with #MultipleSclerosis. Visits to the website are down because those papers don't show up as often. #MSchat
Reposted by Bruno Amaral
Sobre a fusão FCT/ANI. Há coisas com as quais concordo mais, outras menos
E há uma com a qual não concordo NADA: a dupla tutela com a economia
Se tb acham q este é um enorme erro, leiam, assinem e partilhem a carta aberta:
www.publico.pt/2025/11/15/c...
Para assinar: docs.google.com/forms/d/e/1F...
E há uma com a qual não concordo NADA: a dupla tutela com a economia
Se tb acham q este é um enorme erro, leiam, assinem e partilhem a carta aberta:
www.publico.pt/2025/11/15/c...
Para assinar: docs.google.com/forms/d/e/1F...
Por uma ciência autónoma, ambiciosa e responsável
Reformas desta escala, quando feitas sem um claro desenho institucional, modelo de governança e plano de transição rigoroso, correm o risco de gerar incerteza, inércia e paralisia administrativa.
www.publico.pt
November 15, 2025 at 1:14 PM
Sobre a fusão FCT/ANI. Há coisas com as quais concordo mais, outras menos
E há uma com a qual não concordo NADA: a dupla tutela com a economia
Se tb acham q este é um enorme erro, leiam, assinem e partilhem a carta aberta:
www.publico.pt/2025/11/15/c...
Para assinar: docs.google.com/forms/d/e/1F...
E há uma com a qual não concordo NADA: a dupla tutela com a economia
Se tb acham q este é um enorme erro, leiam, assinem e partilhem a carta aberta:
www.publico.pt/2025/11/15/c...
Para assinar: docs.google.com/forms/d/e/1F...
Human-aligned AI models prove more robust and reliable
LLMs perform better when we give them a visual representation on how humans classify things.
#ai #alignet #google #llm #open-science #open-source #to-read
LLMs perform better when we give them a visual representation on how humans classify things.
#ai #alignet #google #llm #open-science #open-source #to-read
Human-aligned AI models prove more robust and reliable
A team from Google DeepMind, Anthropic, and several German partners has introduced a method that helps AI models better mirror how people judge what they see. Their Nature study finds that AI models aligned with human perception are more robust, generalize better, and make fewer errors.
the-decoder.com
November 15, 2025 at 6:37 PM
Human-aligned AI models prove more robust and reliable
LLMs perform better when we give them a visual representation on how humans classify things.
#ai #alignet #google #llm #open-science #open-source #to-read
LLMs perform better when we give them a visual representation on how humans classify things.
#ai #alignet #google #llm #open-science #open-source #to-read
Lucid Voice - Offline Voice to Text for Mac
Offline voice transcription for macOS. Turn speech into perfect text 3x faster with privacy-first AI. No internet required, no subscriptions.
#ai #apple #audio #macos #tools #transcribe
Offline voice transcription for macOS. Turn speech into perfect text 3x faster with privacy-first AI. No internet required, no subscriptions.
#ai #apple #audio #macos #tools #transcribe
Lucid Voice - Offline Voice to Text for Mac
Offline voice transcription for macOS. Turn speech into perfect text 3x faster with privacy-first AI. No internet required, no subscriptions.
www.lucidvoice.app
November 14, 2025 at 6:37 PM
Lucid Voice - Offline Voice to Text for Mac
Offline voice transcription for macOS. Turn speech into perfect text 3x faster with privacy-first AI. No internet required, no subscriptions.
#ai #apple #audio #macos #tools #transcribe
Offline voice transcription for macOS. Turn speech into perfect text 3x faster with privacy-first AI. No internet required, no subscriptions.
#ai #apple #audio #macos #tools #transcribe
Pronto, agora é que ninguém lê Os Maias.
Spotify ganha resumos AI para audiobooks
Spotify ganha resumos AI para audiobooks
O Spotify está a testar uma funcionalidade "recap" para audiobooks, que faz um resumo AI das partes já ouvidas.
ift.tt
November 14, 2025 at 4:15 PM
Pronto, agora é que ninguém lê Os Maias.
Cada vez gosto mais do LinkedIn pelas razões erradas. A normalização do discurso vazio é a suposta sabedoria quotidiana. Se não os podes vencer, diverte-te!
November 14, 2025 at 10:31 AM
Cada vez gosto mais do LinkedIn pelas razões erradas. A normalização do discurso vazio é a suposta sabedoria quotidiana. Se não os podes vencer, diverte-te!
"Do no evil" used to be their motto. www.404media.co/google-has-c...
Google Has Chosen a Side in Trump's Mass Deportation Effort
Google is hosting a CBP app that uses facial recognition to identify immigrants, while simultaneously removing apps that report the location of ICE officials because Google sees ICE as a vulnerable gr...
www.404media.co
November 13, 2025 at 4:27 PM
"Do no evil" used to be their motto. www.404media.co/google-has-c...
Reposted by Bruno Amaral
When I submitted a paper to AI & Society the email confirmation sent me to a website that told me my language quality score was only 3/10 but RUBRIQ could fix it. I study this stuff, so uploaded my paper to RUBRIQ, which proceeded to introduce scientific fabrication and change scientific terms.
November 11, 2025 at 10:56 AM
When I submitted a paper to AI & Society the email confirmation sent me to a website that told me my language quality score was only 3/10 but RUBRIQ could fix it. I study this stuff, so uploaded my paper to RUBRIQ, which proceeded to introduce scientific fabrication and change scientific terms.
November 11, 2025 at 8:13 AM
Reposted by Bruno Amaral
Web summit is coming to Lisbon and I just want it to go away.
November 7, 2025 at 6:08 PM
Web summit is coming to Lisbon and I just want it to go away.
Mas afinal quem foi a pessoa que ontem não dormiu pq esteve a contar os 30mil raios que nos atingiram?
November 5, 2025 at 10:26 PM
Mas afinal quem foi a pessoa que ontem não dormiu pq esteve a contar os 30mil raios que nos atingiram?
Reposted by Bruno Amaral
Remember, remember...
Meet us on Wednesday, November 5th at Cerveteca Lisboa for warm smiles and cold drinks!
#twittlis begins at 18h and ends when the last person leaves.
Join us for a drink and good banter! 😁
www.openstreetmap.org/node/1004078... #lisbon #portugal #meetup
Meet us on Wednesday, November 5th at Cerveteca Lisboa for warm smiles and cold drinks!
#twittlis begins at 18h and ends when the last person leaves.
Join us for a drink and good banter! 😁
www.openstreetmap.org/node/1004078... #lisbon #portugal #meetup
Node: Cerveteca Lisboa (10040782917)
OpenStreetMap is a map of the world, created by people like you and free to use under an open license.
www.openstreetmap.org
November 3, 2025 at 4:57 PM
Remember, remember...
Meet us on Wednesday, November 5th at Cerveteca Lisboa for warm smiles and cold drinks!
#twittlis begins at 18h and ends when the last person leaves.
Join us for a drink and good banter! 😁
www.openstreetmap.org/node/1004078... #lisbon #portugal #meetup
Meet us on Wednesday, November 5th at Cerveteca Lisboa for warm smiles and cold drinks!
#twittlis begins at 18h and ends when the last person leaves.
Join us for a drink and good banter! 😁
www.openstreetmap.org/node/1004078... #lisbon #portugal #meetup
I have been using Claude more often now, and there's quite a difference in culture between it and ChatGPT. Claude feels friendlier and more transparent, and it guides me to concrete tasks. ChatGPT feels like a white box of mystery.
November 5, 2025 at 10:56 AM
I have been using Claude more often now, and there's quite a difference in culture between it and ChatGPT. Claude feels friendlier and more transparent, and it guides me to concrete tasks. ChatGPT feels like a white box of mystery.
As part of my role in the #ai for higher #education project, I wrote up a tutorial with how I am currently using AI tools to prepare classes. ai-hed.eu/post/ai-acad...
It is also my response to those who are pushing back the adoption of AI tools.
It is also my response to those who are pushing back the adoption of AI tools.
Putting AI to Academic Use
A practical guide to integrating AI into academic work without losing agency. Learn how to build a sustainable research workflow using tools like Zotero, Obsidian, and specialized AI search engines—pr...
ai-hed.eu
November 4, 2025 at 8:26 PM
As part of my role in the #ai for higher #education project, I wrote up a tutorial with how I am currently using AI tools to prepare classes. ai-hed.eu/post/ai-acad...
It is also my response to those who are pushing back the adoption of AI tools.
It is also my response to those who are pushing back the adoption of AI tools.
Terno não deve ser fato. Não há nada de terno em usar um fato, não há ternura. E fato não devia ser facto porque fica sem afecto ao qual está afecto.
November 3, 2025 at 2:26 PM
Terno não deve ser fato. Não há nada de terno em usar um fato, não há ternura. E fato não devia ser facto porque fica sem afecto ao qual está afecto.