Google will win the AI race

Do not index

Three years ago, Google declared a ‘code red’ when ChatGPT launched.

Last month, OpenAI issued their own ‘code red’ about Google.

Complete role reversal in 36 months.

Here's how Google turned it around, and 6 practical lessons you can take from it 👇🏼

The crisis (Dec 2022 - Feb 2023)

ChatGPT hits 1 million users in 5 days on launch.

Google panics and rushes Bard to market.

The demo fails publicly - Bard gives a wrong answer.

Stock drops 8%. $100B wiped out in a day.

The irony in all of this is Google invented transformers (The OG AI folks) in 2017.

All eight original AI researchers had already left.

Everyone declared Google dead.

"Too slow, too bureaucratic, game over."

The fix (April 2023)

Google merges DeepMind and Google Brain into one team under Demis Hassabis.

Hundreds of engineers move from Search to AI.

A rival AI lab source says: "Decisions got made faster. They moved faster, and the technology quality f*cking increased dramatically."

That's the move that turned everything around.

The giant was awakened.

The domination (2024-2025)

Gemini 1.5 Pro (Feb 2024):

2 million token context. OpenAI had 128K, Google's was 16x bigger.

NotebookLM (Sept 2024):

Audio overviews go viral on TikTok. 2M+ users overnight.

Deep Research (Dec 2024):

Ships 2 months before OpenAI's version.

Gemini 3 (Nov 2025):

Salesforce CEO: "After 2 hours with Gemini 3, not going back to ChatGPT. The leap is insane."

Sam Altman issues his own "code red" about Google.

This is how Google came back with a better trajectory.

The unfair advantages

Google had many things that others didn’t, and using them might have been the biggest pivotal point for Gemini.

1 - Distribution:

1.8B Gmail users, 2.5B YouTube users, 3.9B Android devices, 91% search share, 65% browser share.

Result: 2 billion using AI Overviews, 650M using Gemini app.

2 - Infrastructure:

10M+ custom TPU chips. Own the full stack. $85B in 2025 CapEx.

Even competitors now pay to use it - Anthropic signed for 1M TPUs, OpenAI runs on Google Cloud.

3 - Capital:

$95B cash, $72B free cash flow.

Meanwhile OpenAI projects $8B losses in 2025. Anthropic relies on Amazon ($8B) + Google (~$2B) funding.

When you have these three, being "first" stops mattering.

Lesson for you:

Use your unfair advantages, because everyone else is using theirs.

What you can steal from this (practical lessons)

1 - Fragmentation kills velocity

Google had competing AI teams. The merger unlocked speed.

→ Your version:

Kill internal competition. If you have two people working on similar problems, merge them or stop one initiative.

Fragmented teams = slow decisions = competitors ship faster.

2 - Infrastructure compounds, hype doesn't

Google spent a decade building TPUs while everyone mocked them for being slow. When they worked, they worked.

→ Your version:

What's your "TPU"? The unsexy foundation nobody sees?

For SaaS: in-house unique data > flashy features

For content: email list > viral tweets

For agencies: SOPs and systems > hero employees

Build infrastructure while competitors chase trends.

3 - Distribution beats innovation

Google didn't build better models.

They used the distribution everyone forgot they had.

→ Your version:

What distribution do you already own but underuse?

Existing customer base for new products

Email list for affiliate offers

Social following for launching MVPs

Past clients for referrals

Build new audiences. But first, squeeze the ones you have first.

4 - Specialization is the only moat against giants

OpenAI's dropping share (76% → 59.5%). Anthropic's rising (12% → 24%) by going developer-focused.

→ Your version:

You can't beat bigger players at their game. Go narrow, talk to your exact customer:

Not "project management tool" → "project management for SaaS founders"

Not "email marketing" → "email for DTC fashion brands with SMS integration"

Not "AI copywriter" → "AI landing page optimizer for SaaS with A/B test built-in"

The riches are in the niches. Especially when giants wake up.

5 - Your biggest asset might be causing paralysis

Google's $175B search business made them scared to ship. That fear almost killed them.

→ Your version:

What's your "search business"?

The $50k/year retainer client stopping you from building products

The consulting work preventing you from scaling

The comfortable salary making you delay your startup

Sometimes your biggest asset is your biggest liability.

Google proved you can have both, but only if you're willing to risk the first for the second.

6 - Speed without structure fails

Google's rushed Bard launch cost them $100B in market cap and credibility.

→ Your version:

Don't confuse "moving fast" with "shipping fundamentally broken things":

Get the MVP out, but not if core functionality breaks

Launch before you're ready, but after it works

Ship weekly, but test internally first

Speed is the goal. But don’t be reckless.

Google's comeback wasn't about building better AI.

It was about finally using assets they already had.

Startups win through speed and focus.

Incumbents win when they stop being scared of their own power.

The real lesson:

First mover advantage is a myth when someone has distribution, infrastructure, and capital you can't match.

Your job isn't to be first.

It's to find the one thing the giant can't or won't do - then own it completely.

Hit reply and tell me: what's the unsexy niche business you're building right now that nobody sees? 👇🏼

Tweet of the week

This is lovely. There will be down days, but it’s important to not lose your calm then.

Read and learn.

That’s it for this week

I will see you next week!

Until then,

Keep building

Tibo 💻

Google will win the AI race

The crisis (Dec 2022 - Feb 2023)

The fix (April 2023)

The domination (2024-2025)

The unfair advantages

What you can steal from this (practical lessons)

Tweet of the week

Join other 65k+ readers now!