Tom Shaughnessy
Founding Partner at Delphi Ventures, Co-Founder of Delphi Digital
Alphabet is repeating its biggest AI mistake by putting resources behind financing Anthropic and selling chip access instead of its DeepMind lab, which is why talent left. In January 2020 Noam Shazeer built a chatbot he said could replace search, but Google sat on a three-year lead to protect search, let him leave, then paid $2.7 billion in 2024 to rehire him. That handed the category to OpenAI and Anthropic, now worth a combined $1.8 trillion. x.com
Google's search is becoming less profitable as AI agents disrupt its core business. Even perfect execution could make the transition painful, and Google is famously poor at new products. Its conduct over the past decade also raises stronger ethical objections than Meta's. Despite strong fundamentals, explosive cloud growth, no real risk and substantial growth ahead, Alphabet is unattractive relative to most other major AI companies. x.com
Alphabet's real AI moat, a few hundred irreplaceable people, is walking out. Jeff Dean, Sanjay Ghemawat, Gemini technical lead Oriol Vinyals and Google Brain co-founder Quoc Le left together as Demis Hassabis stepped back from running DeepMind, six weeks after John Jumper left for Anthropic. Google is a founding investor in their new company and supplies its first-year compute, buying a seat outside after failing to keep them. Infrastructure can be rented; that talent cannot be rebuilt. x.com
Alphabet's search monopoly faces a different competitive game as agent companies build their own search tools. Google Cloud grew 103%, but that was not enough to offset disappointing ads growth. Its only path is to accelerate AI training and keep spending on capex. Search will never be the same, and the risk-reward has changed, though Alphabet remains an optimistic prospect. x.com x.com x.com
The race to win AI is keeping Alphabet on a reckless spending path despite the damage to its share price. The expectation that spending would moderate has been wrong for a year; winning still takes precedence over the financial consequences, and a course correction has yet to arrive. x.com
Alphabet's businesses beyond search are why it will lead the real AI boom. Google Cloud, DeepMind and Waymo broaden its reach, and its $900 million SpaceX investment is now worth $85 billion. Fear emerging as the shares fall is an opportunity; a decline toward the buy zone is welcome, even as Q3 capex has taken a large hit out of free cash flow. x.com x.com x.com x.com
Alphabet's AI buildout is hurting near-term operating profit as capex becomes depreciation. Operating income missed by about 6% and margin fell to about 32% from 34%, while much of the roughly 30% EPS miss came from gains and losses on investments, not Search, YouTube or Cloud. AI remains bullish, but the spending is not good for Google in the short term; the $195-205bn 2026 capex range is unchanged and 2027 capex is expected to rise significantly. Alphabet remains Overweight with a $425 target. … Show more
Alphabet's Q3 search miss strengthens the concern that AI is not great for search. FinancialJuice reports Search & Other revenue of $63.35bn versus $64.85bn expected and capex of $51.36bn versus $55.78bn expected. Anthropic's third quarter also looks less than excellent on this read-through. x.com
Alphabet's search advertising is under attack as chatbot queries grow without matching ad revenue. Google blamed AI Mode on its earnings call. Search ads grew 12% and missed estimates, down from 19% in Q1 and 17% in Q2. Sentiment has completely flipped bearish as its search business shows signs of being eaten by AI. x.com x.com
Alphabet is the most complete vehicle for betting on AI, and it is undervalued. x.com
Alphabet's search business will inevitably be cannibalized by AI, and search is its core advertising revenue driver. Google can pivot, but Meta will always have social networks people depend on to feel connected, making Meta the easier choice. x.com
