🇦🇺 Fantastic to see Australian entrepreneurs & investors building local AI that reflects Australian values, norms, culture and standards. I recently wrote about the compelling reasons to do this, but there's a crucial debate we need to have: 🤔 The Big Questions: 1️⃣ Should Australia Build LLMs? Some industry leaders like Vicki Brady and Scott Farquhar argue that: Australia doesn't need to build its own LLMs The world doesn't need any more LLMs 2️⃣ How Would We Actually Do It? 🏗️ What does "build" really mean? Train from scratch? Adapt existing LLMs through emerging methods? 🛡️ What does "Sovereign" actually require? Must ALL training and inference happen on Australian-owned hardware? Does this approach hamstring development? ⚡ The Innovation Trade-off: Do pragmatic organizations out-innovate those fixated on pure on-premises sovereign AI development? 💡 Key Insight from Global Customers: Building LLMs with on-premises GPUs is incredibly challenging – you spend more time managing hardware than focusing on data and innovation. More insights coming in my upcoming article on how AWS customers around globe are using cloud computing to build sovereign models Thoughts? I'd love to hear from Simon Kriss, Anton van den Hengel, and Dr Troy Neilson on this. 📖 Read my recent deep dive: Building Australia's Sovereign AI Future https://lnkd.in/gvEQ7g9b #AustralianAI #SovereignAI #TechInnovation #AI #Australia #LLMs Tim Bradley Simon Elisha
🇦🇺 Thrilled to share this piece in the The Australian Financial Review by Tess Bennett about Sovereign Australia AI and the journey we're embarking on to build Australis - Australia's own foundational language model. I couldn’t be prouder to be building this alongside my mate and Co-Founder Simon Kriss This isn't just about building another AI model. It's about ensuring Australia maintains its digital sovereignty and voice in an increasingly AI-driven world. We shouldn't be dependent on decisions made in Washington, Silicon Valley, Beijing or Europe about how AI understands and represents our Australian culture, values, and way of life. Yet to achieve this isn’t easy and it takes scale to compete, so we’ve made a massive investment in compute and have acquired 256 NVIDIA Blackwell B200 GPUs hosted in NEXTDC's secure Australian data centre in Melbourne, and managed by our colleagues at SHARON AI What makes me particularly proud of what we’ve done so far: ✅ Ethical from the ground up - We're committing $10M to compensate copyright owners, working WITH creators rather than against them ✅ Built for Australia, by Australians - From the ground up, built, trained and inferenced onshore, here in Australia, and compliant with our privacy and copyright laws ✅ Customers & Partnerships In Place - We partnering with some amazing Aussie organisations to help them solve some of Australia’s AI challenges. The partnerships we can speak about include ACS (Australian Computer Society), UNSW Canberra & GT Systems Australia, with more to follow! The path to digital sovereignty doesn't require billions, it requires vision, ethical principles, and the determination to ensure Australia's voice isn't lost in the global AI conversation. We're proving that sovereign AI can be built responsibly, affordably, and with respect for the creators whose work makes it possible. Australia's digital future should be in Australian hands. Let's build it together. 🚀 Special thanks to Craig Scroggie, Andrew Leece, Kieran Habojan, Dan Mons, Sudarshan Ramachandran, Rhod Brown, Brett Bonser, Melissa Hamilton, Josh Griggs, Rhett Sampson and so many more. #ai #sovereignai #llm #gpt #innovation #australia #australian https://lnkd.in/gnF8N3mR
Simon Kriss
7mo
Hey Craig - Lods of questions here. My short form answers are: 1) We MUST build our own LLM. We risk being dictated to by the US in terms of what we can and cannot have otherwise. Plus we MUST understand what was built and HOW it was built. 2) Build, for us, means a ground up approach - truly foundational. It must be trained on locally held GPUs (to avoid the US CLOUD Act) based in an Australian owned and controlled data centre. We will be operating off a private cloud for our GPUs - provided by an Australian owned and controlled company that has no servers in the USA. Sovereign all the way from the bottom to the top of the stack (with exception of the GPU manufacture as there is no Aussie alternative). Hope this helps.
Glassbox Labs•6K followers
7moAdding to my mate Simon’s comments, and addressing the comments around managing the hardware. This is something we’ve considered deeply, and discussed with all our partners. With that, we’ve built a deep relationship with SHARON AI who’s entire business is built around both managing and scaling AI hardware, here, onshore and absolutely Australian. This doesn’t need to be a binary discussion, and if some industry leader claims they have a bead on the AI market, I’d be seriously questioning their leadership. What I can say, is that, with the best partners in Australia, and a willingness to get off our ass and build this, we’re doing all we can to solve the challenges we face in terms of sovereignty and security. Seriously though, should we listen to the opinions of the largest customers of the companies who don’t want to see sovereign AI succeed in Australia, considering so much of their bottom line is connected to selling the products that compete with sovereign interests? I know I certainly will not be… Feel free to have a chat if you’d like to chat further Craig Lawton.