Full disclosure before we get into this: I'm a Google shareholder. Not just a casual "I bought a few shares" kind of shareholder. When the whole "nobody uses Google anymore, everyone switched to ChatGPT" panic was in full swing and the stock was tanking, I doubled down. While people were writing Google's obituary, I was buying more.

Why? Because the "Google is dead" narrative never made sense to me. Even if they were lagging behind on AI at that point, I couldn't think of a single company better positioned to win this race. Think about the assets for a second. Google Search, Gmail, Google Cloud, YouTube, Google Maps, Android, Waymo. That's not just data. That's the most proprietary, hard-to-replicate data moat on the planet. And unlike every AI startup out there, Google doesn't need to "win" users. People are already living inside their apps. They just need to make those apps smarter.

On top of that, they have essentially infinite money, some of the best talent in the world, their own data centers, and their own custom silicon with TPUs. If this were a strategy game, Google would be the player who spawned with every resource on the map. The only question was whether they'd actually use them.

But resources alone weren't enough for me. I needed to see that something had shifted inside the organization. That at least in some areas, they could operate more like a lean "just ship it" startup rather than the bureaucratic dinosaur they're famous for being. And honestly? I started to feel it. Following people from Google DeepMind on X, reading Logan Kilpatrick's posts - there was a different energy. Less corporate speak, more builder mentality. That was my validation. "Ok, they're ready."

And they delivered. When Gemini 3 dropped, it was the best model across all benchmarks. The same people who were screaming "Google is doomed" suddenly couldn't stop talking about how bullish they were. Funny how that works.

So what's the problem?

A lot has genuinely changed inside Google. They're moving faster, shipping more, and clearly taking AI seriously. But some things? Some things are so painfully, stereotypically "big corporate Google" that it drives me crazy. Both as a user and as someone who literally owns a piece of this company.

Let's start with the one that hurts me the most: design and UI.

Google's interfaces are ugly. I'm sorry, but there's no other way to put it. They're unpolished, inconsistent, and genuinely unpleasant to use. I'm someone who cares deeply about this stuff. I love products like Linear and Resend - tools where every pixel feels intentional, where opening the dashboard is almost a joyful experience. You feel like the people who built it actually cared about how it feels, not just how it works.

With Google products, it's the opposite. I want to open it, get what I need, and close it as fast as possible. There's zero delight. Zero "wow, this is nice." Just utility wrapped in mediocrity.

It's not just how it looks. It doesn't work well either.

I feel like every fifth query I send to Gemini comes back with some cryptic "Unknown error (1)". That's it. No context, no explanation. Just... error. Sometimes Gemini responds with "Sorry, I'm a text-based model, I can't help you with that" to completely normal questions like how long you should fry a steak. A text-based model can't answer a text question about cooking? What?

And my personal favorite: when you ask it to "suggest 5 prompts I can use for image generation," instead of giving you the text prompts, it immediately generates five images. You asked for prompts. It gave you the outputs. That's not a feature, that's a misunderstanding.

Look, these kinds of rough edges would be totally forgivable from a scrappy startup or early-stage Anthropic UIs. But this is Google. The company with more engineering talent per square meter than probably any organization in history. There's no excuse.

The fragmentation problem

Why is there no native Gemini app for macOS? In 2026? When AI assistants are becoming the primary way people interact with computers?

And more fundamentally - why do Google Search with AI Mode, Gemini, and AI Studio all exist as separate things where I can basically do the same stuff? Why isn't this unified into something coherent? It feels like three different teams built three different products and nobody in the room asked "wait, shouldn't this just be one thing?"

The announcement problem

This one really gets me. I love when a company announces something and I can go try it right away. The excitement is there, the momentum is there, the hype is there. You ride the wave.

Google does the opposite. "Here's this incredible new thing! But you can't use it yet. We'll roll it out to Pro users first, then everyone else. When? We don't know. Oh, and it's not available in Europe."

By the time it actually reaches you, you've already forgotten what it was called. Speaking of which - what was Google's AI video generation tool called? The one where you could "vibe code" apps? How many people know about Google Whisk? Or Disco? Or Stitch? Or Jules? Is Jules still a thing or has it been quietly abandoned? Some of these feel like they exist purely for keynote demos, not for actual humans to use.

These are not rhetorical questions. They perfectly summarize the chaos of Google's AI ecosystem. When OpenAI or Anthropic announce something new, everyone knows exactly where to find it and what it does. With Google, it's a treasure hunt where half the treasure might not exist anymore.

The "20 departments" problem

Where do I even manage all of this? Google Cloud? Google AI Studio? Google Workspace? How do I get API keys for something that should be three clicks away? Why does Gemini 3.1 in Antigravity keep crashing or rate-limiting users who are paying $200 a month for Google AI Ultra?

It's painfully obvious what's happening. Twenty different departments are working on things that should feel like one polished product. They keep reorganizing, renaming, reshuffling. The result is a fragmented mess where even power users can't keep track of what's what.

Can this actually be fixed?

Is it easy to solve? Absolutely not. This is a deeply structural problem baked into how a 180,000-person company operates. But I keep thinking - what if Google found their Jony Ive? Someone who could become the face of a new design era at Google. A person with the authority and taste to say "no, this isn't good enough" and actually have that mean something.

On a pure UI and design level, the gap isn't that hard to close. The talent is there. The resources are obviously there. What's missing is a unified vision and the willingness to treat design as a first-class priority, not an afterthought.

I know there are smart people inside Google who are painfully aware of everything I've described here. That's what makes it so frustrating. The potential being wasted is staggering.

I genuinely wish them all the best. But the honest truth is that the things I'm talking about probably won't change without some pretty fundamental shifts inside the company. The kind of shifts that are much harder than building a better model.

Google has every piece of the puzzle. They just need to figure out how to put them together.