I’m starting to think AI models will matter way less than we think

So recently I watched the speech by Genspark CEO at AGI Playground 2026, and one point has been stuck in my head. Models are probably going to become commodities. There are already too many of them, they’re getting closer in capability, and the cost keeps dropping. And honestly, as a normal user, I really don’t want to think about whether this particular task should go to Claude, GPT, Gemini, or whatever model drops next week. I just want the work done. Which made me think the more interesting battle might actually be happening one layer above the models: who has your context? Like your emails, decks, docs, spreadsheets, meetings, research, previous projects, the way you work, the decisions you made months ago. Right now all of that is scattered across 20 different apps. So every time I open a new AI tool, I basically have to introduce myself again. But imagine most of that gradually living inside one AI workspace. The more work it does with you, the more context it accumulates. And the more context it has, the less explaining you have to do next time. That feels like a much stronger flywheel than simply having access to the “best” model. Maybe the real endgame of the all-in-one AI workspace isn’t all your tools in one place. It’s all your context in one place. Just like Genspark. Maybe a lot of people realized this way before I did, and I’m just catching up. Is it just me, or does this seem like where the real AI competition is heading? Curious what you guys think. No judgment please :)