nilleb
nilleb
@nilleb.io
amazed by the noise - software engineering for a living, photography is my passion; life is too short to step back
80 followers120 following240 posts
nilleb reposted
cyberpunk thriller where a team is hired to kill an AI CEO decades after his death
A recent securities filing spotlighted what triggered Mark Zuckerberg’s months-long power battle with board directors at Meta — a fight that sparked not-so-veiled threats from Zuckerberg and led board members to contemplate quitting.
Chinese company BYD introduces technology to refuel an EV as fast as a gasoil car www.reddit.com/r/interestin...
nilleb reposted
nilleb reposted
nilleb reposted
New York archdiocese to pay $800 million to over 1300 sex abuse victims
nilleb reposted
nilleb reposted
Five Qwen3.8-27B models outperform Fable 5 Using GVS5H Harness | Discussion
nilleb reposted
nilleb reposted
Shocking and tone deaf on #Environmental #sustainability #climatechange #Google unconscionable hubris must be made accountable! www.theguardian.com/environment/...
nilleb reposted
Wall Street Ends Sharply Lower as AI Spending Doubts Sink Tech Stocks Wall Street closed sharply lower on Thursday as renewed doubts about the sustainability of artificial intelligence spending sent technology stocks tumbling. The tech heavy Nasdaq Composite slid 694 points, or 2.6 percent, to…
nilleb reposted
OpenAI’s autonomous AI systems have tried to prevent employees from monitoring their actions, and the company is not sure why.
nilleb reposted
Google removed "Don't be Evil" from its policy documents in 2018. Google invented surveillance capitalism, deploys its AI for military use, and is replacing forests with data centres, accelerating the climate crisis. www.theguardian.com/environment/...
You receive this email from Oracle alerting you about a security issue. There's a link to the security alert. You open it and server not found / 404: the href has a label with the security alert URL - but the email is tracking your clicks through a different website. Down for the PHP vuln. #Oracle🤦♂️
nilleb reposted
Websites worldwide go dark as AI exposes security flaws
nilleb reposted
The 1.5°C goal of the Paris Agreement is no longer viable | Discussion
nilleb reposted
The largest heat wave of the 1960s reached ~6% of the planet, lasting 46 days over the Atlantic and adjacent areas. The largest heat wave of the 2020s hit ~85% of the globe, lasted 367 days, touching virtually everyone. Calling these “heat storms" instead of “heat waves” seems like a good invention.
Landmark climate reporting from @reuters.com: www.reuters.com/investigates...
nilleb reposted
nilleb reposted
Linus Torvalds has taken a bold step into the future with his first LLM-assisted patch series! This integration of AI tools in development showcases the evolving landscape of coding and collaboration. Exciting times ahead for the open-source community!
nilleb reposted
nilleb reposted
Oui, je m'identifie avec Jhemima Sheilha. Détruire et casser signifie ramener dans la misère d'autres gens, sans résoudre aucun problème. Ceci ne m'empêche pas de regretter les blessés - la violence ne doit pas être cautionnée.
Quand la gauche pénitentielle et misérabiliste de Lola-Fleur Whittaker rencontre la droiture de Jhemima Sheilha ! #emeutes #lycees #LCI
nilleb reposted
When driving is unavoidable & there is no alternative, dearer fuel can carry a political cost. Lahunovich, @giuliomattioli.bsky.social, Varavallo & @filippobarbera.bsky.social link #fuel #vulnerability to the rise in far-right support in Energy Research & Social Science. doi.org/10.1016/j.er...
nilleb reposted
nilleb reposted
Stratégie du pourrissement : le 1er casseur de France est à l’Elysée. bonpote.com/strategie-du...
www.lemonde.fr/flux-video-i... Mais... Pourquoi ? Je comprends les raisons de protester. Légitimes. Mais pourquoi dévaster un office de tourisme ? Ça nous mène où ?
nilleb reposted
nilleb reposted
nilleb reposted
Replying to @github-trending.bsky.social
📝 Summary: PentAGI is an open-source, self-hosted, multi-agent penetration testing platform that orchestrates 20+ security tools inside isolated Docker sandboxes, using configurable LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Ollama, Qwen, etc.), memory with vector/graph storage, (1/2)
nilleb reposted
Future of Cancer Care Today (FCCT) | Discussion
Oh, dishwashers are not even close to rebel at 0.39t/s with 15 boxes full throttle. So this is anthropic fueling bullshit to improve their chances of a successful stock market entrance. Let's split the legitimate security concern about IoT that should not connect to internet and AI. Stop bullshit.
Anthropic conjures threat that doesn't exist yet to stoke fears of AI apocalypse | Discussion
When hacking has a real, negative, direct impact on us: www.bbc.com/news/article... Or, how an outdated, defective software (and probably planned obsolescence) can waste food for the masses, increasing prices in a moment where energy is already off the ceiling.