I’ve been using it since last fall so I have enough time with it to say something useful beyond the initial impressions.
What it does well: the differentiated reading generator is genuinely good. I paste in a passage and get readable versions at multiple levels that I can use directly or edit minimally. The IEP accommodation suggestion tool is useful as a starting point, though I always modify significantly. The rubric generator saves time at the draft stage.
Where it falls short: anything that requires knowing my specific students. Lesson plans are generic and need heavy customization before they’re usable for my class. The quiz question quality is hit or miss depending on how specific the input text is. The feedback templates are a framework at best.
The honest summary is that it works best as a drafting accelerator for things I already know how to do well, and is much less useful for things that require knowing my classroom context. That’s probably true of every AI tool in this space, but it’s worth naming explicitly so you go in with accurate expectations.
Worth the cost? For me, yes, specifically because of the differentiation tools. For someone who doesn’t need that, maybe not.
the differentiated reading thing from the student side: when it’s good it actually helps me understand stuff i’d otherwise skip because it’s too dense. when it’s bad it’s just a shorter version of the same confusing thing. so quality really does vary
The IEP accommodation starting point use case is interesting because that’s a context where the stakes of a wrong output are high and the need for professional judgment is significant. Using it as a starting point you substantially modify is the right approach. Using it as a finished document would be a different matter.
The ‘drafting accelerator for things I already know how to do well’ framing is the most honest description of AI tool value I’ve seen in a while. It applies well beyond teaching. The tool amplifies existing competence. It doesn’t substitute for it.
A six-month review is more valuable than an initial impression for exactly the reasons you’re describing. The novelty effect wears off and what you’re left with is what the tool actually does when integrated into a real workflow. This kind of long-view honest assessment is what the space is missing.
The differentiated reading generator finding matches mine. That specific feature is more useful than the broader platform for my actual workflow. The lesson plan generation I also found too generic to use without significant rework, to the point where I stopped using it.