Big Tech Copies Startups in Every AI Field – Except Note-taking

• Eric Wu • Product

OpenAI disrupted search. Cursor revolutionized coding. Midjourney redefined design.
From code generation to image creation, almost every hot AI tool in the past three years has followed the same pattern: startups lead, Big Tech scrambles to catch up.

But there's one glaring exception: AI note-taking.
Here, Google's NotebookLM sparked a wave of AI transformation in traditional note-taking software
and paved the way for a new generation of AI-powered note-taking startups.

To understand this anomaly,
we need to look at the broader AI startup trends of recent years.

From ChatGPT to Windsurf

We're all familiar with ChatGPT's timeline by now:

  • November 30, 2022: ChatGPT launches
  • December 4, 2022: Hits 1 million users in just 5 days
  • February 1, 2023: ChatGPT Plus subscription launches

Three months after launch, OpenAI (still considered a startup back then – nobody calls them that anymore, right?)
had already rolled out a paid plan.

So when did Big Tech enter the chat?

The timing and execution tell the story: these products were rushed out in panic.
And just as Big Tech thought they could catch their breath:

  • March 1, 2023: GPT-3.5 API released
  • March 14, 2023: GPT-4 launches with multimodal support

Two weeks. That's all it took to leave Big Tech in the dust again.
The pattern was clear: OpenAI sprinted ahead while giants struggled to keep up.

The same story played out in image generation

OpenAI kicked things off with DALL-E 1 on January 5, 2021.
But the real industry benchmark came with DALL-E 2 on April 6, 2022,
and Midjourney's Discord launch on July 12, 2022.

Big Tech? Late to the party, as usual:

  • May 2022: Google launches Imagen 2
  • March 2023: Adobe Firefly enters public beta
  • March 2023: Microsoft's Bing Image Creator (still powered by OpenAI's DALL-E)

None of these made a dent in Midjourney's market dominance.
The infamous "ChatGPT Ghibli incident" wouldn't happen until March 2025.

The pattern repeated in code generation

OpenAI's Codex API launched in August 2021.
GitHub Copilot officially released on June 21, 2022.
Both predated ChatGPT.

They had users, sure, but not market dominance.

Then Cursor launched its AI IDE in March 2023.
Same familiar VS Code interface, but with deep AI integration.
(GitHub Copilot at the time? Just a chatbot bolted onto VS Code.)
With zero marketing budget, Cursor hit over a million users.
Big Tech finally got the memo: developers don't want a chatbot for looking things up – they want an agent that writes code.

Amazon scrambled to launch CodeWhisperer
on April 13, 2023 – just one month after Cursor.

Google's Gemini Code Assist for individuals? Not until February 25, 2025.
OpenAI's Agentic Codex? May 16, 2025.

None of them could compete with Cursor.

Cursor's real competition came from fellow startups: Windsurf (launched November 13, 2024)
and Anthropic's Claude Code (released alongside Claude 3.5 Sonnet on February 24, 2025).
Both gained rapid market acceptance.
OpenAI and Google even tried to acquire Windsurf.

AI browsers? Same playbook.

Brave's Leo kicked off with Summarizer on March 2, 2023.
Then came Arc (July 25, 2024),
Dia from the same company (June 11, 2025),
and Perplexity Comet (July 9, 2025).

Google, owner of Chrome? Scattered experimental features.
Microsoft with Edge? Launched a Copilot sidebar on September 26, 2023 that nobody uses.
Copilot Mode finally arrived July 28, 2025 – as a limited preview.

NotebookLM: Big Tech's Revenge

Across every AI category, we see the same pattern: startups dominate, Big Tech plays catch-up.
Except for one.

Before NotebookLM, there were scattered attempts to add AI to note-taking apps:

  • Logseq got a GPT-3 plugin in June 2022
  • Roam added an AI Extension in October 2022
  • Obsidian launched an AI Assistant plugin in May 2023

But these were just plugins.
Deeper integrations included:

  • Notion AI (February 2023)
  • Tana (May 2023)
  • OpenAI-backed Mem.ai

Everyone was basically using AI for the same thing: summarization.

Then NotebookLM changed everything.
Launched July 12, 2023 (remember when it was called Project Tailwind?),
it was designed specifically for deep research and document insights.

Suddenly, everyone realized AI wasn't just for summarizing –
it could create and research.
For once, Big Tech led the way.

Why Note-taking Was Different

Note-taking has fundamental differences from other AI applications:

Notes are long-term investments

Unlike chatbots or image generators that produce one-off outputs, notes accumulate over time.
Nobody wants to store years of knowledge on a startup that might disappear tomorrow. Google's reputation solved this trust problem instantly.

Integration with existing data

Google owns Google Drive, where users already store years of documents.
This meant users could instantly bring their existing knowledge into NotebookLM.

The podcast innovation

Alphabet owns YouTube – the world's largest audio-visual database.
This gave Google unmatched advantages in training multimodal models.

The prototype team? Just 4-5 people who built it in two weeks.
They combined startup agility with Big Tech resources (NotebookLM is still free).
These advantages let NotebookLM leap ahead of the market.

Meanwhile, existing startups were stuck:

  • Obsidian was built on local Markdown files – the opposite of cloud-based AI
  • Roam's graph database lacked the semantic layer needed for AI
  • Notion focused on organizing and archiving, but NotebookLM said that doesn't matter – what matters is creating content (NotebookLM's filing system is terrible, but nobody cares)

Big Tech's own note apps were equally stuck:

  • Microsoft's OneNote and Loop had confused positioning – where do you even add AI? (Unless you start fresh like Google)
  • Apple Notes was trapped in a closed ecosystem with no cross-platform integration, and Apple Intelligence was nowhere in sight

Google quickly connected NotebookLM to Google One and Google Workspace,
capturing both consumer and business users.
This helped NotebookLM establish early market leadership.

By late 2024, it entered hockey stick growth, with monthly visits up over 300%.
Shades of ChatGPT's early days.

But Startups Weren't Sitting Still

An investor once told me the golden rule of startups: "Drive nails"

If you see Big Tech doing something, run the other way.

(Nobody's going to invest if you claim to be the next Microsoft 365 or YouTube, right?)

But Google spent millions validating and educating the market.
People realized: "Note apps aren't just for filing – they can help me create content".
Suddenly, startups built from scratch for AI workflows had their opening.

Granola uses AI for meeting notes.
Launched May 2024 (ten months after NotebookLM),
they raised a $20M Series A just five months later,
followed by a $43M Series B in December 2024.

  • Jellypod: Boasts up to four hosts, targeting NotebookLM's two-host limit
  • Wondercraft: Positions itself as a professional-quality AI audio studio
  • Podcastle: Emphasizes human recording + AI editing

These directly target NotebookLM's AI podcast feature.

  • Saner.AI: Productivity tools specifically for ADHD users, recommending past notes and generating plans
  • Tanka: Integrates all team conversations to auto-generate emails, presentations, and investor updates
  • Biolevate: Focuses on medical and healthcare writing, compliance checking

AI output is expensive (output tokens typically cost 3-5x more than input tokens).
Before NotebookLM, startups couldn't afford to validate the hypothesis that "users want output, not just organization."
They also needed to educate the market: "In the AI era, you need to change your workflow."

NotebookLM did both for them.

AI as the Great Unbundler

NotebookLM is like Excel – the spreadsheet giant that startups unbundled piece by piece.

  • Airtable unbundled spreadsheets, adding relational database capabilities without Excel's Office bundle requirement
  • Notion went further, adding document editing and project management to the database core
  • Salesforce unbundled most aggressively, building a complete business platform with end-to-end workflows and industry-specific solutions, turning Excel processes into SaaS

But this time, unbundling is happening faster.
Not just because we've seen the playbook before, but because Gen AI is accelerating everything.

Now Is the Perfect Time to Enter the AI Note-Taking Market

  1. Google has educated the market, accelerating growth

In 2023, the note-taking market was valued at $7.3 billion with a modest 5.4% CAGR.
By 2024, it reached $7.9 billion with CAGR jumping to 16%.

  1. Emerging startups show vertical markets remain underserved

Note-taking spans industries and use cases:
From student lecture notes and researcher literature reviews to corporate meeting minutes and professional knowledge bases.
Every industry has unique note-taking needs.

Despite many players, the market is too broad – gaps remain everywhere.

  1. Note-taking is shifting from "recording" to "creating"

Traditional notes were personal knowledge repositories.
With AI, notes become content generators.
Legacy products become baggage (think Kodak, Blockbuster).
Unburdened startups can move faster.

Why Daino Can Ride This Wave

Thanks to social media, 70% of consumers are influenced by KOLs in their purchasing decisions.
Micro-influencers (10K-100K followers, 47.3% of all KOLs) have 60% higher engagement than mega-influencers.
KOL marketing is growing at 12.5% CAGR – a viable market opportunity.

This massive group of content creators who need consistent output
perfectly aligns with the AI note-taking market's shift to "creation."

With AI-powered note-taking tools,
Daino helps KOLs rapidly transform inspiration into high-quality social content.
Think auto-generated post drafts, audience feedback analysis, and multi-platform publishing.

Daino captures both the AI note-taking market's shift from "recording" to "creating"
and precisely targets KOLs – high-influence content producers –
serving two emerging markets simultaneously.

No dominant player exists in this vertical yet.
Daino's unburdened innovation model
positions us to lead as the content generation platform of choice for KOLs.