CODE IN TOKYO

Why Does Google Not Feel Like It Is Winning the AI Market? One Commentator's Explanation Made Sense to Me

A Question About Google’s Position in AI

The other day, I was watching a video by a commentator discussing the AI industry.

In the video, he talked about a question I had also been wondering about: why does Google not feel like it is winning the AI market, even though it has invested so much in AI?

Of course, this is not because Google’s technical ability is weak.

Gemini is genuinely smart. It is strong in images and video. Google also has Search, YouTube, Android, and cloud infrastructure.

By any normal standard, Google is extremely strong.

And yet, from the perspective of someone like me who writes software with AI every day, the current main character feels like Claude Code. OpenAI is also chasing aggressively with Codex.

So what is Google doing?

To summarize the commentator’s argument rather roughly, the explanation was this:

Google may be aiming at something much bigger.

Programming Is One of the Clearest AI Markets Right Now

At the moment, programming is one of the clearest areas where generative AI can make money.

I use Claude Code every day.

I now let AI handle a large part of my development work, from code generation to testing, debugging, and refactoring.

Corporate software development also involves large budgets. If AI can enter the programming workflow, it can directly capture part of that corporate development spending.

That is why Anthropic is strengthening Claude Code so aggressively, and why OpenAI is putting serious effort into Codex.

To put it bluntly, the rules of competition are very easy to understand.

Which AI can take over more human work?

The more an AI can replace human labor, the more companies will pay for it.

Was Google Aiming at Something Bigger?

This brings us back to the commentator’s point.

In his view, Google is not trying to win only in one area such as programming.

If you look at Gemini, its coverage is extremely broad: text, images, video, audio, and programming.

The idea was that Google may be aiming not for an AI that is strong at one specific job, but for something closer to AGI in the long run.

Of course, this is not something Google has officially said in exactly those terms. It is only the commentator’s analysis.

Still, when I heard it, it made sense to me.

When I use Claude Code, it is very clear what Anthropic wants to do. It wants the AI to write programs.

Recent models can read code, infer intent, investigate the project, and continue working even when my instructions are somewhat rough.

Gemini, on the other hand, can do many things.

But if I look only at programming, which is my actual work, I do not currently feel that I should stop using Claude Code and move to Gemini.

Maybe this is the structure of the market:

While Google was building an all-around player, specialists were taking the market.

Google Has Started Moving

That said, there is one important addition to this story.

Google is not ignoring the programming market.

In 2025, it announced an AI development environment called Antigravity, and it has continued to put effort into it.

Its AI agents do not only write code. They can also operate an editor, terminal, and browser to move work forward.

It is clearly targeting the same market as Claude Code and Codex.

In other words, Google probably does not believe that building a general-purpose AI is enough by itself.

If the commentator’s analysis is correct, Google has finally started looking seriously at the specialized market that is making money right now.

However, at least around me, I still do not hear the name Antigravity nearly as often as Claude Code.

I do not use it myself either.

That gap is significant.

AI tools are not chosen only by performance.

Once a tool enters a development workflow, people do not switch easily.

I have also structured my projects and documentation around Claude Code. Even if Gemini improves somewhat, I do not feel like throwing that away and moving everything to another tool.

Google may not have been late in technology so much as late in trying to capture a specialized market.

Was Google’s Bet Wrong?

Of course, it is too early to say that Google has lost.

If Google really creates something close to AGI before everyone else, the current competition could be overturned completely.

The categories themselves, such as programming AI, video generation AI, and image generation AI, might disappear. We might simply use the smartest AI for everything.

Anthropic and OpenAI are aggressively going after markets that can make money right now.

Google may have been aiming at something bigger.

And judging from Antigravity, Google has now also started going after the market directly in front of it.

That is what I felt after hearing the commentator’s argument and looking at the actual products.

I do not know whether this view is correct.

But today, I am still writing programs with Claude Code. Not Antigravity.

I do not know who will win the future AGI race, but at least this month’s $100 is being paid to Anthropic.

Just because Google released Antigravity does not mean my $100 will immediately move to Google.

That may be what it means to capture a specialized market first.

Contact

I provide system development and technical consulting using AI, AWS, and Claude Code.

Contact form: https://holly-money-e94.notion.site/390ff30cf18c8086a676fe630d171873