AI at Meta’s Post

I have developed the base code of an advanced AGI, I have been trying for days to find support because I want to establish myself in Switzerland to establish the basis of this AGI. I even sent emails to people in the government, But since I didn't know who to contact, they simply told me that "they don't deal with those matters" and that they "wished me luck." That's why here I am, staying up all night (in my country it's dawn when in Switzerland it's daytime, there's a 9-hour difference) writing.a roadmap, doing market research, reviewing budgets, quoting flights, investigating VC's and their emails or numbers, spending on phone credit to do so. I'll be back in Switzerland soon, and I'll make it anyway. I just want you to know that what matters more to me than recognition is caring for the nations and people.who are not taken into account in the technological race of the titans, competing for money, to be the first, without looking at the many who are left behind. The ethical code of my AGI is NOT FOR SALE. Whoever wants to support it must accept that I lead, and I lead to balance this scale that the powerful have tipped in their favor. I will be the voice of those who do not are heard and I will be the forcethatwillmakeadifence

Excited for this

The Llama Defenders Program is an impressive initiative aimed at addressing AI-enabled security risks. The collaboration with CrowdStrike for CyberSOC Eval and the AutoPatchBench for automatic patching of vulnerabilities shows a comprehensive approach to enhancing security. Seeing companies like Zendesk already using these tools to detect AI-generated voices is impressive and highlights their real-world impact.

Suka ini

Love this

Sounds really insightful

Llama Guard, Llama Firewall, Llama Prompt Guard… if my next security breach isn’t stopped by an alpaca in tactical gear, I’m going to be very disappointed. 

Love seeing AI security get the spotlight! Llama Guard 4 and Llama Firewall look like huge steps toward building safer multimodal systems. Excited for how Prompt Guard 2 versions balance performance and speed, much needed for real-world deployment! 

🦙🔐 Meta just raised the bar for AI security at #LlamaCon. With new open-source tools like Llama Guard 4 and Llama Firewall, developers can now protect their AI systems from prompt injection, insecure code, and risky plugin behavior—right out of the box. 💡 And there’s more: Prompt Guard 2: Faster, cheaper jailbreak and injection detection CyberSecEval 4: Real-world benchmarks for AI resilience Llama Defender Program: Early access for trusted partners to cutting-edge AI defenses 🔐 As AI capabilities grow, so must our security strategies—and Meta’s tools are a big leap forward. 🛡️ Building your own AI chatbot or assistant? Security isn’t optional—it’s foundational. With BotGPT you can: ✅ Train on your own content (PDFs or websites) ✅ Deploy in minutes—no code required ✅ Maintain full control over privacy and output 👉 Start now → bot.wordgptpro.com Bring secure, custom AI to your business today.

It's inspiring to see advancements in AI security like these.🔐 At qantum.one/ai-infrastructure, we're firm believers in the power of secure AI systems. We help companies deploy and scale their private AI systems, and these types of tools are absolutely essential in ensuring not just functionality, but more importantly, safety. Besides this, our Bionic Testing approach at qantum.one/qa-automation also combines human intellect with AI to deliver efficient QA Automation services. Looking forward to more innovations in this space! 👏 #AIInfrastructure #SecureAI #AI #QAAutomation #qantumone