The Real AI Shopping War Isn't About AI. It's About Data.

Salutations, Olio aficionados! 👋

Welcome to the 262nd edition of Weekly Olio. We’re thrilled to introduce a fresh new twist to your Sundays: Publisher Parmesan, our hand-picked, thoughtfully crafted edition designed to spark inspiration and insights for the week ahead.

It’s the perfect way to unwind, recharge, and prepare for the week with something truly worth savoring.

If you’re new here and you’re looking for more long-form, crispy writing, click the link to subscribe under this GIF 👇

Instead of hiring developers, marketers, and dozens of tools just to get started, Willo helps people turn business ideas into actual online businesses IN MINUTES without requiring technical knowledge.

It's for people who've had ideas sitting in their notes for months but never knew how to move forward. They're also offering Founding Member pricing at 50% OFF for a limited number of early users. That "maybe someday" business idea? This might finally be the thing that gets it out into the world.

For the past two years, the AI industry has been obsessed with models. Which company has the smartest AI? Which model scores highest on benchmarks? Which chatbot can reason better, write faster, or generate more convincing content?

Those questions matter. But they may not be the questions that determine who wins the next phase of AI.

A more important battle is beginning to emerge—one that has less to do with model intelligence and more to do with something far harder to replicate: proprietary data and real-world distribution.

Alibaba's recent integration of Qwen and Taobao offers a glimpse into what that future might look like. On the surface, it appears to be another AI shopping assistant. In reality, it may represent something much bigger: the emergence of data as the primary moat in the AI era.

For investors, founders, and technology leaders, this shift is worth paying attention to. Because if the first chapter of AI was about building intelligence, the second chapter may be about owning the environments where intelligence becomes useful.

The idea of an AI that shops on your behalf is hardly new. Science fiction imagined it decades ago. In Iron Man, Jarvis could place orders with a few spoken commands. In Her, an AI assistant handled everyday tasks through conversation. Even when Amazon launched Echo in 2014, Jeff Bezos spoke openly about a future where shopping could happen through a simple voice interaction.

The vision always seemed obvious.

Humans spend countless hours searching for products, comparing prices, reading reviews, and making purchasing decisions. Why wouldn't software eventually automate that process?

The answer, as it turns out, was that understanding human intent is much harder than most people imagined. For nearly a decade, the technology simply wasn't ready. Today, that is finally changing.

Looking back, AI shopping has gone through three distinct phases.

The first was the voice-assistant era.

Products like Amazon Alexa and Google Shopping Actions allowed consumers to place orders using voice commands. The idea sounded revolutionary. The reality was far more limited. These systems worked only when users already knew exactly what they wanted.

"Buy paper towels." "Order more toothpaste." "Purchase batteries."

The moment shopping became exploratory rather than transactional, the experience broke down. The systems could identify keywords but could not understand intent.

The second phase emerged around conversational customer service.

Companies like Taobao, JD.com, and Amazon deployed chat-based assistants. But these systems weren't really shopping advisors. They were customer support tools disguised as intelligence.

They could answer questions about shipping, returns, and order status. They could not meaningfully help consumers decide what to buy.

The third phase began with large language models.

For the first time, machines could understand context rather than keywords. That changed everything. Suddenly, a user could ask:

"I'm looking for a lightweight travel backpack for a two-week trip through Europe, preferably under $150."

And the system could understand what the user actually meant. The bottleneck shifted from understanding language to completing transactions.

As AI capabilities improved, the world's largest technology companies arrived at three very different conclusions about how AI shopping should work.

The first strategy was pursued by model companies. OpenAI became the clearest example.

Last year, ChatGPT launched Instant Checkout through partnerships with Shopify, allowing users to complete purchases directly within conversations. It seemed like a logical extension of ChatGPT's growing role as a product discovery tool.

Then something interesting happened.

Users preferred researching products through ChatGPT rather than buying through it. Merchants were slow to integrate. Adoption lagged expectations. By early 2026, OpenAI reportedly began scaling back the initiative.

The second strategy came from e-commerce incumbents.

Amazon launched Rufus directly inside its marketplace. Unlike OpenAI, Amazon already controlled inventory, fulfillment, payments, reviews, and logistics. The AI assistant could leverage the entire ecosystem.

The results were impressive. Amazon reported that Rufus accumulated hundreds of millions of users and meaningfully increased purchase likelihood. Yet challenges remained. Critics pointed to limitations outside Amazon's internal ecosystem and questions around recommendation quality.

Then there is the third strategy. And this may prove to be the most powerful one.

Alibaba is attempting something neither OpenAI nor Amazon can easily replicate. It owns both the model and the marketplace.

The integration between Alibaba's Qwen and Taobao is not simply an AI assistant sitting next to a shopping platform. It is a deep integration where the model can actively participate throughout the transaction journey.

A user can discover products, compare alternatives, place orders, track deliveries, and manage returns through a conversational interface.

This sounds like a small difference. It isn't.

Historically, most AI systems have acted as advisors. Alibaba is trying to turn AI into an operator. That distinction matters because consumers ultimately care less about recommendations and more about outcomes.

The value is not finding a product. The value is successfully buying it.

With AdQuick, executing Out Of Home campaigns is as easy as running digital ads. Plan, deploy, and measure your real-world advertising effortlessly — so your team can scale campaigns and maximize impact without the headaches.

Visit AdQuick.com

Most investors still think of AI competition as a model race. Bigger models. More parameters. Better benchmarks. That view is becoming increasingly incomplete.

As foundation models mature, performance differences between leading systems are narrowing. Intelligence is becoming abundant. What remains scarce is data.

This is where Alibaba's position becomes particularly interesting.

Taobao has accumulated more than two decades of shopping behavior and billions of product listings. Every search query, comparison, purchase decision, return request, and customer interaction creates valuable training data.

That data teaches AI something far more important than language. It teaches behavior.

Knowing which product a user ultimately buys after comparing ten alternatives is incredibly valuable information. Knowing what causes customers to abandon purchases is valuable. Knowing when users prefer convenience over price is valuable.

These are not language problems. They are decision-making problems. And decision-making data may become one of the most valuable assets in the AI economy.

For much of the last decade, technology companies built moats through algorithms. The AI era may look different. Increasingly, competitive advantages appear to be shifting toward proprietary data loops.

Every interaction improves the system. Every transaction generates more training data. Every successful recommendation strengthens future recommendations. This creates a feedback loop that becomes difficult to replicate.

A competitor can copy a feature. They cannot easily copy twenty years of customer behavior. This is why many experienced investors are beginning to view data ecosystems as more durable than model advantages.

Models can be replicated. Data often cannot.

Chinese venture capitalist Zhu Xiaohu captured this idea well when he argued that once model capabilities stabilize, competition shifts toward engineering execution and closed data loops. That shift is already happening.

The deeper story here extends beyond shopping. Alibaba is not merely integrating Qwen into Taobao. It is gradually connecting AI across its broader ecosystem, including payments, travel, mapping, local commerce, and logistics.

Competitors are pursuing similar strategies.

ByteDance is linking AI with Douyin Commerce. JD.com is building AI shopping capabilities. Meituan is exploring similar territory.

What looks like an AI competition is increasingly becoming an ecosystem competition. The winning company may not be the one with the smartest model. It may be the one that owns the richest collection of real-world actions.

Whenever a new technological platform emerges, investors initially focus on the technology itself. Eventually, attention shifts toward distribution. Then it shifts again toward data.

We may be entering that third phase now. The most important question is no longer: "Which company has the best AI model?"

The more important question is: "Which company owns the highest-quality feedback loop?"

Alibaba's Qwen-Taobao integration offers a glimpse of what that future might look like. Not because conversational shopping is guaranteed to succeed. But because it reveals where durable value may ultimately accumulate.

If consumers become comfortable shopping through conversations, the winners will not simply be the companies with smart AI. They will be the companies sitting on the largest reservoirs of behavioral data. And those reservoirs are much harder to build than another model.

The history of technology is full of moments when investors focused on the wrong layer of the stack.

During the internet era, many people focused on websites while infrastructure quietly became enormously valuable.

During the mobile era, attention centered on apps while platforms accumulated the real power.

AI may be entering a similar phase.

The first chapter was about intelligence.

The second chapter may be about ownership of data, workflows, and real-world behavior.

If that's true, then the most valuable AI companies of the next decade may not be the ones with the smartest models.

They may be the ones with the deepest understanding of how billions of people make decisions.

Interested in learning more about AI? Check out our previous coverage here:

Salary is only part of the equation. Local taxes, benefits, and employment costs add up quickly.

Use Oyster’s calculator to estimate the real cost of hiring globally, in seconds.

Estimate hiring costs

That’s all for this week. If you enjoyed this edition, we’d really appreciate if you shared it with a friend, family member or colleague.

We’ll be back in your inbox 2 PM IST next Sunday. Till then, have a productive week!

Peeyush and Mohit

Disclaimer: The views, thoughts, and opinions expressed in the text belong solely to the author, and not necessarily to the author's employer, organization, committee or other group or individual.