Laurent Bourrelly

Nikky aka DJ Rituala
The witch with turntables

youtube.com/@underground...

www.undergroundtechnogirl.com

Nikky aka DJ Rituala
The witch with turntables

youtube.com/@underground...

www.undergroundtechnogirl.com

Jumping around Sunday

Human Made

My favorite tech of 2026 is my #iPod collection

To ChatGPT : Create a caricature of me and my job based on everything you know about me.

Cheeky monkey

Sprint

Rodeo

All good

She knows

Perfect day

Horsing around

I see you

This new feature in Google Gemini changes everything forever.

Visual Layout makes ChatGPT look ancient.
It redefines the world of Search and nobody was ready.

Good stuff 👇👇👇

OpenAI Database of 220 gold tasks huggingface.co/datasets/opena…

Research summary by OpenAI (easy read, what they did, overall results): openai.com/index/gdpval/

The actual research paper (to dig into the details and breakdown by profession):
arxiv.org/pdf/2510.04374

Alphabet is increasingly launching 'moonshot' projects as independent companies -- here's why | TechCrunch Despite the emphasis on detachment from ideas, X employees do have significant skin in the game when projects spin out.

Alphabet is increasingly launching ‘moonshot’ projects as independent companies

techcrunch.com/2025/11/02/a...

Forget about Plan B.

That’s the only secret to achieve massive success in life.

Unless…

Garage cleanup

I’m stumbling upon some serious gear.

Retr0bright treatment and it’s good to go

My mindset is low tech back to vintage tech at the moment.

Fuck the Cloud

Told you so

If you wait until China « wakes up, » you lost.

China already won.

In the meantime you are just a customer of Silicon Valley.

We are free (from money, censorship, technological limitations, etc,) with Open Weight and Open Source LLMs.

Chinese LLM are better all around.

Incredible AI for cybersecurity… or… hacking portswigger.net/burp/ai

Sora Transform text and images into immersive videos. Animate stories, visualize ideas, and bring your concepts to life.

Who is on Sora 2 ?

sora.chatgpt.com/profile/laur...

First #Sora2 creation

The future of AI is now www.kimi.com

Finally, AI automation is becoming less dumb than the past two years.

www.getneutron.com

How AI and Wikipedia have sent vulnerable languages into a doom spiral Machine translators have made it easier than ever to create error-plagued Wikipedia articles in obscure languages. What happens when AI models get trained on junk pages?

How AI and Wikipedia have sent vulnerable languages into a doom spiral

www.technologyreview.com/2025/09/25/1...

Call Me A Jerk: Persuading AI to Comply with Objectionable Requests

Call Me A Jerk: Persuading AI to Comply with Objectionable Requests

gail.wharton.upenn.edu/research-and...

kimi-k2 A state-of-the-art mixture-of-experts (MoE) language model. Kimi K2-Instruct-0905 demonstrates significant improvements in performance on public benchmarks and real-world coding agent tasks.

Are you just a customer, using Claude or another subscription based LLM for coding?

Stay free

ollama.com/library/kimi-k2

PS : Open Source from China is superior to paid Silicon Valley

YouTube video by Bg2 Pod NVIDIA: OpenAI, Future of Compute, and the American Dream | BG2 w/ Bill Gurley and Brad Gerstner

NVIDIA CEO Jensen Huang about OpenAI, Future of Compute, and the American Dream youtu.be/pE6sw_E9Gh0?...

Stress Testing Deliberative Alignment for Anti-Scheming Training — Apollo Research Future AIs might secretly pursue unintended goals — “scheme”. In a collaboration with OpenAI, we tested a training method to reduce existing versions of such behavior. We see major improvements, but ...

Pay attention 👇👇👇

Stress Testing Deliberative Alignment for Anti-Scheming Training

www.apolloresearch.ai/research/str...