It’s been quiet the past few months, and I thought some of you had abandoned ship. The reality is completely different. I haven’t stopped interacting with the AI - quite the opposite. I’ve had so much fun writing about it that there’s more to write about.
Work behind the silence
Since September, I’ve been building a strategy to implement AI across our trust. It wasn’t part of any official role expansion, it was just something that needed to be done when someone needed attention, no harm done. The work was intensive, both effective strategies and unexpected problems were discovered. There is much to discover from this experience, and subsequent posts will explore what I learned about literacy and integration at scale.
What tickles me is that it mirrors what I’ve seen many times in education: people dealing with emergence are often stretched between doing the work and documenting it. Both are rare.
A great quarter for AI education
Looking back from September to early December, I couldn’t agree more because I started writing about this industry. Not necessarily muscular announcements or top-of-the-line releases, but also the convergence of practical educational tools that reach their true maturity.
Consider what has happened these past few months. Gemini has seen significant upgrades. ChatGPT has evolved significantly with a real education version. Google Gemini has features for the classroom environment. The LM notebook continued its evolution. Microsoft, after what felt like a period of small talk, is starting to make more sense if its luxe suite of educational software. Their partnership with Anthropic for Electric Copilot is a real step forward, especially for those working in Microsoft’s dominant education environments.
And then there’s an open mine of upgrades that they’ll properly explore their capabilities. The fact that we remain committed to providing free access to educators around the world makes these events especially important.
The task of keeping pace
The truth is that the pace of development has really taken a hit. Not in the “it’s very interesting” way, but not in the “we’ll evaluate and implement it with some responsibility” way. Every update, every new feature, every platform evolution needs careful consideration before bringing it into an educational context. We need to understand not only what these tools can do, but also what we need to do in our specific settings.
This creates a tension between being current and thoughtful. Between documentation developments and implements them in practice. Between sharing and protecting the time needed to develop these insights in the first place.
Moving forward
I will try to participate in weekly posts from this point on. Not the burst of multiple posts that characterized my previous approach, but also the steady rhythm that acknowledges the work in education while trying to keep up with rapidly evolving technology.
Instant messages include the solid work I’ve done, progress logs, and scale logs as well. What strategies have worked for different stages of digital confidence when pitching to colleagues? How do you keep yourself open for opportunities? What does meaningful integration of AI look like as it moves beyond pilot projects and individual classrooms?
A brief apology
To those who have followed this blog from the early days, I apologize for my silence. I know consistency, if you’ve ever felt like a moving target, not being me didn’t help you navigate these waters. The irony wasn’t lost on me that I had too much to work on when it came to AI work.
But maybe it’s appropriate. This field requires practitioners, not just commentators. It needs people willing to test these tools in real classrooms with students with real limitations. Writing problems, but only for practice.
What has changed
Looking back on these months, it feels real. We’ve mostly moved from experimental AI tools that worked in an educational context to AI tools that work. The distinction is very important. We are no longer trying to redesign consumer products for classroom use. We work with platforms that have educational use cases built into their design from the ground up.
This does not mean that these tools are perfect or that their implementation is straightforward. This conversation is about “could it ever work in education?” “How are we good at this job?” This shift represents real progress, as far as possible.
Road ahead
Normality is restored here – this, this, this, at least this, but normality feels like the most difficult concept in educational technology. Maybe that’s the point. We are not going back to normal. We find a new rhythm that recognizes the pace of technological change and the enduring realities of educational practice.
The ride continues. Let’s see where it takes us.