2025 was a year where a lot of assumptions across the tech industry were seriously put to the test. Instead of simply making predictions or tallying wins and losses, I wanted to write out a handful of takes on what actually happened and what it says about where things are headed next.

1. OpenAI cannot get distracted by building too large a fleet of products.

I was extraordinarily excited when Sam Altman announced that he was teaming up with Jony Ive to build hardware back in May. But much of that excitement has settled across the industry as it’s becoming less clear what kind of shape it will actually take. Personally, I think their ambitions will be peeled back and we’ll end up with some sort of phone or earbuds. Not new form factors. But this isn’t just about hardware, which to be fair I do think stands a very real chance of succeeding when it ultimately hits the market, it’s about the software. I fear OpenAI may get distracted by over-productizing their technology. While I personally use Atlas, I don’t know that it’s important enough to make a substantive impact. The average user (and even some techies) don’t quite see the value in AI browsers. The same goes for apps like Sora and the upcoming erotica offering.

OpenAI does its best work when it focuses on the models and the core app, which I fear it is getting slightly distracted away from. It’s no secret that GPT 5 didn’t quite live up to the hype and subsequent models have largely been playing catch up to others. Though I still believe GPT 5 was a fantastic release, the future belongs to whoever has the best model and spending too much time on products at this critical time when they’re arguably still in startup mode is risky. Not to mention they ought to leave some room for a proper ecosystem of third-party products built using their technology to flourish instead of trying to supplant them.

2. Google’s dominance will continue well into the future after climbing back into the top spot.

Just a few years ago folks were shouting from the rooftops that Google was doomed. I may have been one of them, but I’m happy to say that I was wrong. This year Google proved it’s still got the juice. AI mode shipped as their answer to Perplexity, the first real search engine challenger. Gemini 3 blew everything else out of the water and combined with nano banana has driven astronomical usage growth. Veo, Flow, NotebookLM, Project Mariner, among other tools have proven their size and dominance has made it easy for them to spread out and build new kinds of products that users can experiment with. Their hardware has hit a real stride with the new Pixels, which not only have some of the best on-device AI features, but are so good that it’s hard to recommend any other Android device to an iPhone switcher.

There’s very little doubt in my mind that Google’s not going to continue down this incredible path through 2026 and you could argue that they’ll be even more dominant than ever once Siri integrates with Gemini as rumored. Their size alone would be an advantage, but they’ve got the right people and the right corporate structure to make all the right things happen.

3. Alan Dye deserves much more credit for his tenure at Apple, whether people at the company liked him or not.

I really like Alan Dye. His stage presence is strong, he’s soft spoken, he’s absurdly creative, and he brings a different kind of vibe to user interface design. I understand the extensive criticism that exploded across the web when Dye announced he was leaving Apple for Meta. But to act as though his 12 year stint leading UI at Apple was a failure is to deny reality. He isn’t just one of the most influential designers in modern history, he’s a great one. He may not have started out as a UI designer, but designers are often multidisciplinary. So many skills can be spread across a variety of domains. Nobody’s tenure is perfect, but all you have to do is look at the facts. iOS 7, watchOS, tvOS, the iPhone X experience, the Big Sur redesign, the Dynamic Island, visionOS, Liquid Glass and so much more have all defined his time at the company. He led a team that made our devices more vibrant, more playful, and arguably more powerful. He tends to be seen as someone who cared more about content than the interface itself but I just don’t quite buy that argument. He helped give our devices a very distinct new identity several times over, making sure they always felt fresh.

I suspect that his time at Meta will bear fruit for Zuck and team. It will be nice to see other products benefit from his exquisite taste. Is it a good thing that Apple will have some fresh leadership? Of course, that’s natural for every company and it only means that the next few years will be even more fun to speculate about. We may see the return of some more classical facets of the Apple ecosystem, something I am particularly excited about. But declaring everything to be immediately better is premature. I know it’s been said that folks inside of Apple that worked under Alan were relieved that he departed the company and are excited about working for Stephen Lemay. But that’s not the point. You have to ask, was his tenure a net positive, was the output beautiful, and did everything just work? The answer to all three of those things is a resounding yes.

4. Software on-demand is already here, Apple and Google are going to need to do more to adapt.

When vibecoding started to become a thing, I was immediately drawn to the idea that software on-demand would be the future of apps. With new offerings like Wabi, canvas tools in ChatGPT, Gemini, and Claude, and services like Lovable I suspect that we are only seeing the very beginning of a complete shift in the way software is built, distributed, and improved upon. I can now prompt any app into existence right from my phone in just a few minutes. Apple and Google both rely on a smaller subset of developers to fill their app stores with software, but what happens when the consumers become the developers? The whole model will have to change. I hope we start to see the first glimpses at how that might happen very soon. I suspect that these kinds of new user-made applications won’t always live inside of a third-party app, but right alongside all of your other content.

5. Apple couldn’t deliver on AI, but it showed off in other ways. That being said, 2026 needs to be big.

This year’s Apple hardware is some of the best they’ve ever shipped. This is especially true when it comes to the iPhone family and the MacBook Air. Liquid Glass and the complete revamp that applied to all platforms made it a fun year for software too. They delivered incredible new services content like Pluribus, The Studio, F1, among others. It was a strong year, but they still couldn’t ship the one thing that they desperately have to. Apple ultimately got away with delaying the new Siri in 2025, but they’re not going to get a pass next year. It needs to be truly spectacular. Maybe a refreshed executive suite will help refocus the company on the things that truly make Apple, Apple.