Approach AI with 5 key questions: Before using AI, ask: - what type of AI - what data has been used to train it - how might the system fail - who benefits from my use - can I test before I buy Watch Rose's full videos on machine learning and baking at https://t.co/foNYwTiBsz

⏰ This time tomorrow: AI is transforming education faster than most governing boards can respond, and without strategic oversight, schools risk fragmented implementation that undermines their values. Book a seat: https://t.co/MkmQEHcLpq

"There are some things are AI systems simply can't learn." Human judgement is irreplaceable and amazing, and AI cannot replicate it effectively because AI is an algorithm disconnected from the whole of the real world.

📁 Yann LeCun says that scaling models will not get us to human intelligence. He explains that the industry remains obsessed with making LLMs bigger, but that this path is fundamentally broken. It does not matter how many parameters we add or how many clusters we build, because

"Tech lobbyists have frequently opposed even the most sensible federal standards and rules — such as... taking down unauthorized content. In an act of total hubris, they are now arguing that states should be banned from regulating A.I." What a disaster. https://t.co/i2WDjCtYS8

The perfect optimised machine learning network. Similar to the making of a choux pastry bun, we must be careful not to overfit our AI system nor give it too little training data. We must tweak the settings just right so it performs in the way we want, otherwise we're cooked!

"Just like with AI, we don't have the knowledge instantly, we have to search for the right settings" Machine learning AI systems need settings and parameters in order to function. Baking these in requires time and money and the balance is why the models can behave that way.

AI systems are locked in place before they're released, and their unique development can lead to 'emergent properties'. This is a bit like baking profiteroles, where the layering of the dough leads to a unique pastry in the oven. Each pastry tastes and looks different.

"Ever wonder why AI is so hugely expensive?" Exponential growth, embodied in these lovely little cinnamon twirl pastries, can explain why. Watch the full video https://t.co/foNYwTiBsz on YouTube

2 DAYS LEFT ⏰ Discover the proven strategic AI framework driving real results - for businesses, small or large, educational institutions and organisations, come learn why early adopters of AI are seeing 40-60% efficiency improvements! #ai #work #data https://t.co/R8RPyd5pqt

Rose Luckin has 30 years of expertise in ethical AI for education and is asked to speak around the world on the future of work. If you're not using AI effectively in your business already, find out from her how to do it. https://t.co/R8RPyd4RAV

If you're looking to push your leadership of AI in schools to the next level, this might a great programme for you... @ASCL_UK present Advanced AI Leadership in Schools - In collaboration with @aiineducation @Knowldgillusion and @CanopyCIC https://t.co/tcsHbKR1Lb

Responsible innovation in digital transformation is non-negotiable - keep ethics front and centre, and learn how AI has the potential to transform the classroom. Learn more => https://t.co/i0zGbFNbQE

Each step was driven by one core goal: ensure that AI tools enhance education without compromising student well-being or teacher autonomy.

We began with a systematic literature review, refined early drafts through teacher focus groups, and then collaborated with a multidisciplinary expert panel.

The motivation was simple: AI holds enormous potential to enrich learning experiences and improve outcomes, yet it also brings significant risks around privacy, bias, and overreliance on automation. The result of our multi-phase process is a practical, adaptable set of controls.

Throughout the last year, our partnership researched existing AI ethics guidelines, conducted in-depth case studies, and engaged directly with teachers, school leaders, and EdTech developers.

Hugely excited to announce a new framework of ethical controls designed to guide the responsible implementation of AI in education. A huge thank you to our partners @Avallain in this endeavour. Why did our two organisations develop this project though? https://t.co/i0zGbFNbQE

🚨 TOMORROW🚨 🎉 Get to grips with AI at AEPA's next roundtable 👩‍💻 Professor Rose Luckin (@Knowldgillusion) will lead the discussion and detail a strategic framework for human-centred AI in education. 🎟️ Get your tickets now: https://t.co/4G5fI5obNB #AEPARoundtable

🎉 Join us for a roundtable focusing on getting to grips with AI 👩‍💻 Professor Rose Luckin, @Knowldgillusion, will lead the discussion, detailing a strategic framework for human-centred AI in education. 🎟️ Get your tickets now: https://t.co/FIiSsqAlFp #AEPARoundtable