Interesting & encouraging result: reasoning models such as o1 become more robust to adversarial attacks as they think for longer.
SECURITY
-
Google Announces Parfait: Private Aggregation & Federated Learning
By
–
Announcing Parfait — private aggregation & retrieval, federated, analytics, inference, & training — a GitHub organization that we’ve developed to demonstrate our state-of-the-art methods across 4 privacy pillars. Learn how it was created & is being used →
https://
goo.gle/4axVNfN -
GenAI Transforming Healthcare: Scaling AI Responsibly and Ethically
By
–
GenAI is transforming healthcare.
From enhancing diagnostics to streamlining operations, GenAI can provide smarter, faster, and more accessible care.
But the real challenge? Scaling AI responsibly with secure, private, and ethical solutions.
Dive into how leaders are shaping the -
AI Grandma Chatbot Traps Scammers in Endless Conversations
By
–
That is a genius use of AI! Meet the AI Grandma who’s every scammer’s worst nightmare.
— Pascal Bornet (@pascal_bornet) 22 janvier 2025
She’s programmed to trap scammers in endless, time-wasting conversations.
In a recent poll, 71% of people said they wanted revenge on scammers—well, this delivers something even better: A… pic.twitter.com/LRHGdi53LWThat is a genius use of AI! Meet the AI Grandma who’s every scammer’s worst nightmare.
She’s programmed to trap scammers in endless, time-wasting conversations. In a recent poll, 71% of people said they wanted revenge on scammers—well, this delivers something even better: A -
Is DeFi Really Decentralized or Just CeFi with Extra Steps
By
–
Correct. Show me this DeFi that you speak of. Mostly I see CeFi, with extra steps.
-
Perplexity Ensures Data Security with New API Launch
By
–
Plus, your data is secure: Perplexity does not conduct LLM training on your data. Start building by generating your API key now:
-
Meta AI’s Content Moderation on Mark Zuckerberg Trump Relationship
By
–
Interesting what Meta AI does & doesn’t say about Mark’s inauguration photo. The unprompted AI-generated summary says comments were mixed, but if you ask Meta AI what they said about Mark’s relationship to Trump it starts writing but auto-deletes. More examples in the thread.
-
Building Defense AI Together with Lecornu
By
–
Merci pour votre confiance @SebLecornu, construisons ensemble l'IA de Défense 🇫🇷 https://t.co/fP9cnbyk6P
— Arthur Mensch (@arthurmensch) 20 janvier 2025Merci pour votre confiance @SebLecornu
, construisons ensemble l'IA de Défense -
EU Commission launches healthcare cybersecurity protection action plan
By
–
Commission unveils action plan to protect the health sector from cyberattacks https://
ec.europa.eu/commission/pre
sscorner/detail/en/ip_25_262
…
#digital #europe #CyberSecurity #healthcare #innovation #technology @BetaMoroney @CurieuxExplorer @Shi4Tech @smaksked @debashis_dutta @nafisalam @Khulood_Almani @enilev -
Algorithm Transparency and Data Protection Regulation Proposal
By
–
I’d make them open the algorithm, restrict data gathering, and commit to first amendment provisions. Regular audits. A law that, even if it were twisted and used against domestic platforms in the future, would be consistent with our constitution.