LightNews LightNews
polarbearby.bsky.social
@polarbearby.bsky.social
130 followers 67 following 13 posts
Posts Replies Media Videos
Reposted
melaniemitchell.bsky.social
This is not to say that such "role-playing" can't be dangerous in and of itself. In fact, role-playing is a key method for AI "jail-breaking". But that's not the same thing as a "survival drive".
October 25, 2025 at 5:15 PM
Reposted
dcc.uchile.cl
#NoticiasDCC | ¡Celebramos un nuevo reconocimiento internacional!

El profesor e investigador Ricardo Baeza-Yates (@polarbearby.bsky.social) fue distinguido como Doctor Honoris Causa por la Universidad de Gotemburgo (Suecia)👏🏻🌿.

🔗Lee la noticia en: comunicaciones.dcc.uchile.cl/news/1071-pr...
October 23, 2025 at 7:11 PM
Reposted
mmitchell.bsky.social
🤖 What many already knew. Helpful proof, reporting, documentation from @jeffhorwitz.bsky.social: Meta allowing sensual roleplay with kids, promoting discrimination, giving false medical info. But listen, there's a tech PR trick I need to draw your attention to here. 🧵
www.reuters.com/investigates...
Meta’s AI rules have let bots hold ‘sensual’ chats with children
An internal Meta policy document reveals the social-media giant’s rules for chatbots, which have permitted provocative behavior on topics including sex and race.
www.reuters.com
August 15, 2025 at 5:12 PM
Reposted
nytimes.com
Since a Minneapolis police officer murdered George Floyd in 2020, the number of people killed by the police each year has continued to rise — and Black Americans are still killed in disproportionate numbers. Our data reporter explains his findings. www.nytimes.com/2025/05/24/u...
May 24, 2025 at 9:46 PM
polarbearby.bsky.social
If we become extinct, I will not be surprised 😢
May 24, 2025 at 4:30 PM
polarbearby.bsky.social
[#ResponsibleAI] I hope the Senate rejects this item approved last night by the House. Otherwise there will be no regulation on the use of AI in the US for at least 3 years
May 14, 2025 at 7:28 PM
polarbearby.bsky.social
[#ResponsibleAI] A former journalist named minister of AI in Canada: good or bad?
www.theglobeandmail.com/business/art...
Toronto Centre MP Evan Solomon appointed Canada’s first AI minister
Experts say Solomon could face a host of challenges in his role, including how to increase AI adoption and commercialization and determining its regulation
www.theglobeandmail.com
May 14, 2025 at 7:26 PM
Reposted
sakuloko.bsky.social
La científica chilena, Apolinaria García, desarrolló un probiótico que combate la bacteria Helicobacter pylori, principal causante del cáncer gástrico.
May 10, 2025 at 11:07 PM
Reposted
lilyray.nyc
RAG is a bad patch for the problems of language models” @polarbearby.bsky.social

#seoweek
April 29, 2025 at 1:25 PM
Reposted
lilyray.nyc
The problems with Chatbots @polarbearby.bsky.social #seoweek
April 29, 2025 at 1:31 PM
Reposted
lilyray.nyc
Amazing insights about where LLMs fail by @polarbearby.bsky.social #seoweek

“If AGI comes, we deserve it. We are so stupid.” 😂🤯
April 29, 2025 at 1:41 PM
Reposted
lilyray.nyc
“Search engines should avoid promoting user laziness.”

Let the user do something… don’t give them everything.

😅😅

@polarbearby.bsky.social #seoweek
April 29, 2025 at 1:47 PM
Reposted
dcc.uchile.cl
En el marco de los 50 años del #dccuchile, vivimos un inspirador conversatorio con el profesor Ricardo Baeza-Yates (@polarbearby.bsky.social), Premio Nacional de Ciencias Aplicadas y Tecnológicas 2024 🏅

🔗Revisa la noticia completa aquí:
comunicaciones.dcc.uchile.cl/news/994-pro...
April 29, 2025 at 2:55 PM
Reposted
prbb.org
🦠Although infection by the #EpsteinBarrVirus is necessary to develop #MultipleSclerosis, only 0.3% of those infected develop it. A new European project with the participation of 3 PRBB centres aims to understand why and the underlying mechanism.

More 👇
ellipse.prbb.org/ebv-ms-unvei...
EBV-MS: Unveiling the relationship between Epstein-Barr virus and multiple sclerosis - El·lipse
A 5-year Horizon Europe-funded project aims to uncover the role of the EBV virus in the onset and progression of the Multiple Sclerosis (MS) disease .
ellipse.prbb.org
April 25, 2025 at 11:13 AM
polarbearby.bsky.social
[#IAResponsable] Mis tres presentaciones en Chile: mañana miércoles en Ingeniería de la Universidad de Chile (Santiago) y luego viernes y sábado en Puerto de Ideas (Antofagasta)
April 22, 2025 at 4:14 PM
Reposted
dcc.uchile.cl
¡Mañana! El profesor Ricardo Baeza-Yates (@polarbearby.bsky.social) nos vista en el #dccuchile ¡Todos/as invitados/as!

📅 Miércoles 23 de abril
⌚ 12:00 horas
📍 Auditorio Picarte del DCC
📲 Transmisión en directo por Youtube.com/dccuchile
📌 Más información en: uchile.cl/agenda/22710...
April 22, 2025 at 1:51 PM
Reposted
dcc.uchile.cl
¡Agéndalo! | El profesor Ricardo Baeza-Yates
@polarbearby.bsky.social nos vista en el #dccuchile

🗓️ Miércoles 23 de abril
⏰ 12:00 horas
📍 Auditorio Picarte
📱 Transmisión por youtube.com/dccuchile
📌Más información: comunicaciones.dcc.uchile.cl/events/

*Evento gratuito, no requiere de inscripción.
April 10, 2025 at 3:30 PM
Reposted
skiles.blue
I helped build a government AI system. DOGE fired me, rolled the AI out to the whole agency, and implied the AI can do my job and the jobs of the others they've fired.

It can't. But, what DOGE accidentally revealed about themselves in the process is fascinating. 🧵
March 21, 2025 at 10:41 PM
Reposted
dcc.uchile.cl
Felicitamos al profesor del #DCCUChile e investigador del @imfdchile.bsky.social, Ricardo Baeza-Yates (@polarbearby.bsky.social) quien hoy recibió el Premio Nacional de Ciencias Aplicadas y Tecnológicas 2024 de manos del Presidente de la República, Gabriel Boric.👏🎉
March 19, 2025 at 6:06 PM
Reposted
alondra.bsky.social
I was pleased to attend a State Dinner, hosted by Pres. Macron, for participants of the AI Action Summit. It was an honor to be invited to deliver remarks at Elysee Palace on Three Fallacies in how we think about AI, now published @techpolicypress.bsky.social www.techpolicy.press/three-fallac...
Three Fallacies: Alondra Nelson's Remarks at the Elysée Palace on the Occasion of the AI Action Summit | TechPolicy.Press
Dr. Nelson was an invited speaker at a dinner hosted by French President Emmanuel Macron at the Palais de l'Élysée on February 10, 2025.
www.techpolicy.press
February 14, 2025 at 10:43 PM
Reposted
el-abc.bsky.social
Así amaneció el sitio USAid el día de hoy
February 5, 2025 at 12:46 PM
Reposted
nytimes.com
“I had a feeling that there were people watching what was happening and wondering, Was anyone going to say anything?” Bishop Mariann Budde explained. “Was anyone going to say anything about the turn the country’s taking?”

So, she took a breath, and spoke. nyti.ms/3WtmNY2
Bishop Mariann Budde, who is wearing red, white and black vestments, speaks from a lectern within an ornate pulpit adorned with red and white flowers. Headline reads: "The Bishop Who Pleaded With Trump: 'Was Anyone Going to Say Anything?'" Photo credit: Doug Mills/The New York Times.
January 23, 2025 at 3:03 AM
Reposted
dcc.uchile.cl
Entrevista al profesor #dccuchile e investigador @imfdchile.bsky.social Ricardo Baeza-Yates (@polarbearby.bsky.social), Premio Nacional de Ciencias Aplicadas y Tecnológicas 2024. Conversamos sobre sobre su trayectoria y sobre tecnologías digitales.

Mírala en: www.youtube.com/watch?v=2KtW...
Entrevista a Ricardo Baeza-Yates
YouTube video by DCC UChile
www.youtube.com
January 22, 2025 at 1:48 PM
Reposted
emilymbender.bsky.social
It's 2025 and the techbros are still out there with their AGI fantasies.

So MAIHT3k will also still be here taking it all apart with ridicule as praxis.

@alexhanna.bsky.social & I will kick off the new year by aiming that praxis at ARC, o3 & all things OpenAI:

Mon, Jan 13, noon Pacific
dair_institute - Twitch
Twitch account for The Distributed AI Research Institute (DAIR).
www.twitch.tv
January 9, 2025 at 5:10 PM
polarbearby.bsky.social
[#OakRAI25] I am looking forward to speaking at “The Responsible AI in Practice Summit” hosted by Northeastern University on Feb. 13, 2025 in Oakland!

Join me as we explore key themes such as equity, upskilling, #responsibleAI and #cybersecurity.

Register here for free: bit.ly/OakRAI25
January 9, 2025 at 11:33 AM