(#112) Has Europe lost the AI battle?; Google’s earnings and the strategy shift
This week, I’m in the USA, and for the first time in a while, I feel like I’m experiencing the “true” internet - a glimpse into the future of digital intelligence and accessibility.
Some of the things that stand out:
? Deep Research – A $200/month Pro plan with 100 monthly queries. Amazing at presenting factual data and handling complex analysis.
? Apple Intelligence – All of it, not just parts.
? Real-time AI copilots – Seamless integration across work and personal life.
? Search that actually works – Less SEO clutter, more precise and useful results.
? Hyper-personalized AI assistants – Tools that anticipate needs, not just react to inputs.
? Instant access to cutting-edge tools – No geo-restrictions, no waiting.
?? It’s a reminder that innovation moves at different speeds depending on where you are. Some of these experiences will take years to reach other parts of the world.
Onto the update:
Amazon’s ongoing struggles with physical retail
1/ Amazon’s online dominance does not guarantee success in the physical world
Despite being the king of e-commerce, Amazon has repeatedly struggled to make its brick-and-mortar experiments work. From Amazon Go to 4-Star and bookstores, none of its retail ventures have scaled successfully?. The failure isn’t due to a lack of capital or technology, it’s a fundamental misalignment between Amazon’s logistics-first mindset and the realities of physical retail, which require customer service, real estate expertise, and local market adaptability.
2/ Amazon is treating its physical stores as a tech incubator, not a retail business
The Amazon Go concept was less about selling snacks and more about developing Just Walk Out technology, which the company now licenses to over 200 retailers ?. Instead of doubling down on its own stores, Amazon has pivoted to selling its tech, a classic platform strategy rather than a retail one. This shift reflects a calculated retreat, where Amazon extracts value from its innovations without the capital-intensive burden of running stores.
3/ Grocery is Amazon’s last big bet in physical retail, but it’s still an uphill battle
With Amazon Go shrinking and its 4-Star stores closing, Amazon’s focus has turned to grocery, a market it entered through Whole Foods and Amazon Fresh?. However, its Fresh stores have struggled to find their niche, even after scrapping Just Walk Out technology in favor of dash carts. The challenge remains: can Amazon win in a market defined by thin margins, local preferences, and established players, or will grocery, like Go, end up as another expensive experiment? LINK
Adobe’s launched AI-powered contract analysis
Three insights from the recent blog release.
1/ Adobe is turning contract confusion into an AI-driven advantage
Contracts are a universal pain point, 70% of consumers sign agreements without understanding all terms. Adobe’s Acrobat AI Assistant is capitalizing on this widespread friction by offering AI-powered contract analysis, making legal documents more accessible and actionable. The move positions Adobe not just as a document tool but as a business intelligence platform.
2/ AI-powered contract analysis could redefine enterprise workflows
By integrating contract summarization, key term extraction, and document comparisons, Adobe is embedding AI deeper into enterprise operations?. This isn’t just about reading contracts faster, it’s about enabling finance, legal, and procurement teams to automate complex reviews, creating a more data-driven approach to legal and business decision-making.
3/ Adobe is taking a safer, enterprise-friendly approach to AI adoption
Unlike competitors that rely on third-party LLMs, Adobe is not training its models on customer data?. This security-first approach makes its AI tools more appealing to regulated industries like finance, healthcare, and law, where data privacy is non-negotiable. In a market where AI trust is still being debated, Adobe is playing the long game by prioritizing corporate adoption over AI hype. LINK
on Apple’s earnings
1/ Apple’s services business is now the real growth engine, hardware is the platform
While iPhone revenue dipped 1% year-over-year, Apple still posted record revenue and profits, driven by a booming Services segment. Services revenue surged 14% YoY to $26.3 billion, making up over 21% of total revenue, but likely contributing a far greater share of profit due to 75%+ margins. This underscores Apple’s evolving strategy: hardware sales create the installed base, but services extract increasing lifetime value. The more devices Apple sells, the stronger its recurring revenue flywheel becomes.
2/ China is no longer a growth market for apple, it’s a fight for survival
Apple’s Greater China revenue declined 11%, marking six straight quarters of year-over-year decline?. This isn’t just cyclical, it’s structural. Apple dominated China’s high-end market when Huawei was crippled by U.S. sanctions, but Huawei’s return with its Mate 60 Pro and domestic 5G chips has flipped the narrative. With Chinese consumers increasingly opting for local brands, Apple’s China growth story is over, the best it can hope for is retention.
领英推荐
3/ Apple’s AI strategy is cautious, but could quietly reshape the iphone experience
Unlike OpenAI or Google, Apple isn’t chasing hype-driven AI announcements. Instead, it’s embedding AI-powered features into the ecosystem, autocorrect, image editing (Genmoji, Clean Up), and writing tools?. Cook positioned Apple Intelligence as a driver of better iPhone 16 sales in AI-supported markets, but China was left out. While AI alone isn’t selling iPhones yet, Apple’s on-device AI approach aligns with its privacy and ecosystem moat, setting the stage for deeper integration down the road.
4/ Finally, although not mentioned in the call, Apple is rumored towork on Mac-connected Augmented Reality glasses. What a pity…and a mess. Earnings, Glasses, Investors
Google’s earnings and the strategy shift
1/ Google Cloud’s growth slowdown isn’t about demand, it’s a supply problem
Google Cloud grew 30% year-over-year, down from 35% the previous quarter, triggering a 7% drop in Alphabet’s stock ?. But the issue isn’t demand, it’s capacity constraints. CFO Anat Ashkenazi confirmed that Google exited the quarter with more AI demand than available capacity, forcing it to ramp up CapEx to $75 billion in 2025 (far above Wall Street’s $58B estimate). This isn’t a weakness, it’s a sign of how intense the AI arms race has become. Google’s challenge is no longer proving AI demand, it’s scaling up fast enough to capture it.
2/ Search isn’t collapsing under AI pressure, it’s holding steady and monetizing well
Despite fears that AI overviews would cannibalize search revenue, Google’s search ad business remains stable, and early results show monetization at approximately the same rate ?. Even more interesting, AI-generated results are increasing search usage, particularly among younger users. While many assumed AI would weaken Google’s search dominance, it’s doing the opposite: driving engagement and keeping users within the ecosystem. If this trend holds, Google’s core revenue stream will remain resilient, buying it time to monetize AI in new ways.
3/ China’s antitrust investigation into Google is more about geopolitics than policy
China’s sudden antitrust probe into Google follows its broader trade war escalation, including tariffs on U.S. imports and restrictions on key semiconductor materials?. Given that Google has no real business in China, the investigation is largely symbolic, a warning shot in a broader economic standoff. But a separate, legitimate antitrust probe into Apple’s App Store policies in China could be far more consequential. If China forces lower app store fees or external payment options, it could set a precedent for regulatory pressure on Apple outside China as well. LINK
Has Europe lost the AI battle?
The article from FT reveals at least three ideas:
1/ Mistral’s efficiency advantage is no longer unique, and China just proved it
Mistral built its reputation on doing more with less, creating cutting-edge AI models without the massive compute budgets of OpenAI or Anthropic ?. But DeepSeek’s breakthrough model from China, built with even fewer resources, has now outperformed Mistral at its own game. This challenges Mistral’s core value proposition and raises existential questions: If efficiency is no longer a differentiator, what is it? The AI race is increasingly about scale, infrastructure, and partnerships, areas where Mistral remains a fraction of its U.S. and Chinese competitors.
2/ Europe’s AI hopes are pinned on mistral, but its funding gap is enormous
Mistral isn’t just a startup, it’s Europe’s only real bet on large language models after Aleph Alpha pivoted away from LLMs. But its $1.2 billion in funding pales in comparison to OpenAI, Anthropic, and now DeepSeek, each backed by war chests 10x bigger. With just 150 employees versus thousands at its rivals, Mistral is in a precarious position: too big to fail quietly, but too small to compete at scale. If Mistral stumbles, Europe may be forced to rely on American or Chinese AI models, undermining its push for technological sovereignty.
3/ Mistral is at risk of becoming an acquisition target, despite its independent stance
Mistral CEO Arthur Mensch insists the company is “not for sale”, but industry insiders suggest a buyout is inevitable. Microsoft already holds a small stake, and Big Tech has been on an AI acquisition spree, absorbing startups like Inflection and Adept. If Mistral struggles to raise more capital, it could face pressure to sell to a U.S. giant or a European conglomerate. The EU’s antitrust regulators may resist a U.S. takeover, but without a deep-pocketed European investor, Mistral’s fate remains uncertain in a rapidly consolidating AI landscape. LINK
Thanks for reading onStrategy! Subscribe for free to receive new posts and support my work. https://plus.onstrategy.eu/
I just launched my book on strategy:
Through 28 chapters I covered three parts: (1) Strategy, (2) Innovation & Growth, and (3) Generative AI.
See a full sample - the chapter on Network Effects. - click HERE