Hi from Tanzania! 🇹🇿
Love Hanoi but the pollution is heavy going this time of year so when the chance came to get on a plane out to East Africa, I jumped at it. If you’ve never been, I couldn’t recommend Tanzania more – the people, the landscapes, the wildlife, the music, the food 😍 #Iliveherenow
But then see what happens when you take a week off? A global race for AI dominance has quietly shifted into overdrive. From $500bn megaprojects to surprising open-source challengers, this month’s developments are reshaping not just how we teach, but who might lead the AI education revolution. Let’s dive in! 🚀
🌐 The Half-Trillion Gambit | Global AI Strategy
Stargate marks a new partnership between a who’s-who of modern computing with some serious financial and governmental backing: OpenAI, Arm, Microsoft, NVIDIA, and Oracle + 500bn from SoftBank with the blessing of the new Trump administration. It may be almost an afterthought in the context of the very next section in this update but it’s still important as it marks what we had thought was our inevitable direction of travel until very recently.
What does this mean in practice? Lots of data centres and jobs in the US but, ultimately, Trump says one key goal is to keep the US ahead of China in the global race for AI supremacy #foreshadowing. One last interesting twist is that First Buddy Elon Musk is apparently sore about this whole thing – he’s been left out of this project and has been trading heat with Altman since. But then maybe all that investment isn’t necessary 🍿👀
🇨🇳 Player 2 Has Joined the Game: China’s AI Revolution
Just when everyone was digesting Stargate’s price tag, DeepSeek R1 crashed the party! Picture this: a Chinese open-source model that cost just $6m to train (vs OpenAI’s $100m+ for GPT4) while matching top-tier performance at 1/30th the cost. Mind-blowing enough? Never to be upstaged, Jack Ma’s Alibaba just dropped Qwen 2.5 on Lunar New Year, claiming even better results! 🎊
The plot thickens: while OpenAI cries IP theft (and it’s very hard not to feel a bit of schadenfreude here), DeepSeek’s clever Mixture of Experts approach suggests raw computing power isn’t the only path forward. Yes, there are content restrictions/biases and major data privacy concerns, but that hasn’t stopped its rocket ride to the top of app stores globally. 📱
Hilarious that AI lost its job … to AI
For educators? JISC’s advice is spot on: explore the offline version yourself (or try Ollama for an easy start with offline AI), but hold off on student recommendations for now. After all, open source AI isn’t going anywhere – might as well get ahead of the curve! 🎓
🔧 Tools of Tomorrow | Platform Evolution
Sam Altman’s week was meant to go so differently! Just as DeepSeek stole his thunder, he unveiled the new Operator agent – imagine an AI that navigates the web just like we do, handling everything from restaurant bookings to grocery runs. While the built-in guardrails are reassuring (it’ll definitely double-check before spending your money on those Warriors tickets! 🏀), watching a computer use a mouse feels a bit like teaching your grandmother to text… surely there’s a more elegant solution? These are computers after all!
The real story though? The giants are feeling the heat from our new Chinese competitors. OpenAI has thrown open the gates to their o1 reasoning model, Microsoft keeps “reinventing” the same features for Copilot, and Anthropic (most writer’s fave AI 📚) just dropped Citations. Need help navigating this explosion of AI tools? The excellent Ethan Mollick has your back with his 2025 Guide to Picking Your AI.
Mollick’s Opinionated Guide is a must-read
🎓 Creativity Unleashed | Learning Innovation
Interesting from ASU’s Punya Mishra and his innovative “Human Creativity x AI in Education” course, where he’s teaching students to view AI not as a replacement for human creativity, but as a powerful amplifier of our innate creative abilities. Through hands-on experiments like rapid-fire visualisation of abstract concepts and guided by Mishra’s seven thinking tools (perceiving, patterning, abstracting, embodied thinking, modelling, playing, and synthesising), students are learning to harness AI as a “smart, drunk, biased intern” – capable of generating countless ideas while leaving the critical thinking and synthesis to human minds.
The course’s weekly student-authored blog posts offer a window into this groundbreaking approach, showcasing how AI can serve as a springboard for deeper creative exploration when combined with human judgement, expertise, and imagination. Will be watching this with interest
🔮 The AGI Horizon | Future Implications
The recent Davos gathering just gave us a glimpse into what might be the most world-shifting tech revolution yet – the potential dawn of AGI. Picture this: some of the brightest minds in AI (we’re talking Yoshua Bengio, Andrew Ng, Thomas Wolf, and Yejin Choi) all in one room, tackling the question that’s keeping tech leaders up at night: are we actually on the brink of AI systems that could outperform humans at, well… almost everything? 🤯
The stakes are astronomical. Anthropic’s Dario Amodei dropped a bombshell that would make even sci-fi writers do a double-take: AI might compress a century of scientific progress into just 5-10 years (and before you dismiss this as tech hyperbole, remember Google’s AlphaFold revolution? 🧬). But it’s not all techno-optimism – the panel dug deep into the thorny questions of embedding ethical values into superintelligent systems and understanding what makes human intelligence unique. With even OpenAI insiders admitting to being “pretty terrified” by the breakneck pace of development, one thing’s clear: the future is arriving faster than anyone predicted! 🌟
Watching these developments unfold – from Stargate’s massive investment to China’s rapid innovations, from evolving tools to creative classrooms – one thing becomes crystal clear: we’re not just witnessing an arms race in AI, we’re seeing a fundamental shift in how knowledge is created, shared, and applied. For those of us in HE, this isn’t just about adopting new tools – it’s about reimagining what learning could look like in a world where AI capabilities are simultaneously more powerful and more accessible than ever before. 🌟
The real question isn’t whether AI will transform education (that ship has sailed! 🚢), but how we’ll shape that transformation. Will we lean into the creative possibilities demonstrated in ASU’s classrooms? Can we balance the promise of tools like DeepSeek with the critical thinking skills our students need? As AI continues its breakneck evolution, perhaps our greatest opportunity lies not in choosing between human and artificial intelligence, but in discovering entirely new ways they can work together.
PS. How beautiful is Tanzania?!? Current view from Nungwi, Zanzibar
