Google Making Billions of AI – The Investment Case for Alphabet

5/5 - (3 votes)
How Google makes $BILLIONS with AI 🀯

Google’s annual revenue jumped from $100 billion to over $300 billion in six years:

YearRevenueGrowth
2011$37,905
2012$46,03921.46%
2013$55,51920.59%
2014$66,00118.88%
2015$74,98913.62%
2016$90,27220.38%
2017$110,85522.80%
2018$136,81923.42%
2019$161,85718.30%
2020$182,52712.77%
2021$257,63741.15%
2022$282,8369.78%
2023$307,3948.68%

πŸ’ˆ Pillar 1: Research Leadership

Google integrates AI teams under Google DeepMind for accelerated AI innovation.

US Base Salary for Robotics Research Scientist $133k to $240k (https://boards.greenhouse.io/deepmind/jobs/5842013):

Google DeepMind Focus on developing Gemini and other AI models for generative applications.
The release of Gemini 1.5 Pro includes breakthroughs in understanding longer complex contexts and multimodal capabilities, as well as advanced NLP.

πŸ’ˆ Pillar 2: Building World-Class AI Infrastructure

Google’s custom TPUs, now in their fifth generation, are crucial for training and deploying AI models like Gemini.

Google’s investment in AI-specific data centers and TPUs optimizes operational efficiencies and scalability.

Infrastructure enables over 100x efficiency improvements in AI operationsβ€”critical for businesses scaling AI deployments.

πŸ’ˆ Pillar 3: AI-Driven Search for Billions of Queries

Integrating generative AI in Google Search, enhancing accuracy and response quality to complex queries.

Experimentation in Search Labs leads to AI overviews on search results pages.

They use the RAG model:

To become an AI engineer and master retrieval augmented generation (RAG) and VertexAI, check out our course here:

πŸ’ˆ Pillar 4: Expanding Global AI Product Footprint

AI features integrated across multiple Google products with wide user bases. Products like Pixel, Photos, Chrome, and Messages enriched with AI capabilities.

🎯 Google Achievements: Six products with over two billion users each, supporting broad AI feature testing and deployment.

πŸ’ˆ Pillar 5: Speed & Efficiency in AI Development

πŸ’ˆ Pillar 6: Monetizing AI Innovations

Cloud and AI infrastructure as a service (1000+ new products in 8 months)

  • AI Hypercomputer + NVIDIA GPUs + TPUs
  • >60% of funded gen AI startups and 90% of gen AI unicorns ($1B+) are Google Cloud customers
  • AI Devs for Vertex AI and other AI APIs

Over 100 million paid Google One subscribers, with new AI-enhanced plans introduced.

AI helps in traditional segments to serve more targeted ads

  • YouTube Ads +21% YoY growth
  • Search Ads +14% YoY growth

Putting It All Together

  • Alphabet remains one of the global AI powerhouses
  • Alphabet revenues still grow in mid-teens (~15%)
  • Alphabet cost of revenue doesn’t grow as fast (~13%)
  • Operating efficiency increases
  • YouTube ads + cloud AI infrastructure are strong
  • Risk factor: Search may be used for RAG-style AI solution so it’s still okay

Valuation of Alphabet Stock (Napkin Math)

Alphabet can always “harness” global attention one way or another (search not needed).

Example: 2 billion people spending 2h/d on {Android phone, Chrome, YouTube, Google Search, AI tools, Photos, …}

  • πŸ‘‰ 1460 billion hours of attention per year
  • Example: $1 monetization per hour (profit)

Using this approach, I put together this spreadsheet with a realistic valuation of $2,920 per share at the time of writing — if they decided to “turn on” their revenue opportunity.