From large models, robots to dating apps: The market's pricing standards for AI in 2026 will fully shift towards return on investment!

As the investment boom in artificial intelligence (AI) enters deeper waters, the narrative logic on Wall Street is undergoing a fundamental shift. According to Morgan Stanley's latest report on the outlook for the internet industry in 2026, the market's pricing standard for AI technology has shifted from mere expectations of technological breakthroughs to a rigorous assessment of return on invested capital (ROIC). The ability to convert computing power into tangible revenue and profit has become the sole measure determining the valuation of tech stocks.

According to the Wind Trading Desk, the analyst team led by Brian Nowak at Morgan Stanley pointed out in the report that the market theme for 2026 will continue the trend of 2025, where capital will concentrate on companies that can demonstrate substantial returns from GenAI or GPU-driven technologies. This means that only companies showing faster revenue growth, higher user engagement, and expanded earnings per share (EPS) and free cash flow (FCF) will receive premium rewards from the market.

Under this new standard, Morgan Stanley is particularly optimistic about giants like Amazon, META, and Doordash, which can achieve efficiency improvements and business expansion through AI, and expects hyperscalers to experience strong growth. In contrast, sectors facing disruptive uncertainties from autonomous driving (AV) or agentic technologies, such as ride-hailing, online travel, and smaller advertising platforms, will receive lower valuation multiples from the market.

The report details the "Top Ten Debates" that will reshape the industry landscape in 2026, covering key topics such as the evolution of large language models (LLM), the application of AI in physical and robotic fields, changes in the search landscape, and the revival of dating apps, painting a comprehensive picture for investors of how AI technology is spreading and monetizing in the real economy.

Debate 1: The "Arms Race" of Large Models Cools Down, Application is Key

In 2026, the parameter competition of cutting-edge models will no longer be the market's excitement point. While we will still see the emergence of new models like Grok 5, Claude 5, and GPT-6, investors' focus will completely shift to productization and monetization.

Google: The key lies in how the Gemini model is deeply integrated into search, YouTube, and cloud services to drive revenue growth.

Meta: The market is watching whether Zuckerberg's "superintelligence lab" can produce SOTA (state-of-the-art) models and convert them into advertising revenue and user stickiness.

Amazon: The growth of AWS is a given, but more importantly, whether AI applications on the retail side (such as the Rufus shopping assistant) can bring in real sales increments.

Debate 2: Where is the ROIC for GenAI?

In 2025, the market was still anxious about AI capital expenditures (Capex); by 2026, the market demands to see returns. Morgan Stanley predicts that 2026 will see a stepwise leap in the adoption rate of GenAI technology on the enterprise side (from single-digit growth to high double digits) Cloud Giants Benefit: Amazon AWS, Google Cloud (GCP), and Microsoft Azure will experience stronger growth than expected. The surge in backlog orders indicates that enterprises are migrating workloads on a large scale to meet AI demands.

Diffusion Effect: As power and computing bottlenecks ease, AI technology will spread from tech giants to broader economic sectors.

Investment Insight: Look for companies with accelerating growth in cloud business backlog orders, as this is a leading indicator of ROIC about to be realized.

Debate Three: Is the Wave of Layoffs Not Over? The Efficiency Dividend Brought by AI

This may be the most brutal yet profit-margin-friendly trend. 2026 could be the first year that tech giants significantly enhance internal efficiency using GenAI, thereby slowing hiring or even continuing layoffs.

Profit Release: Morgan Stanley's model assumes that the growth rate of non-depreciation/non-advertising operating expenses (mainly labor costs) for Meta, Amazon, and Google will significantly decline.

Valuation Support: If these companies can control labor costs while growing revenue, there will be substantial upside potential for EPS (earnings per share) and FCF (free cash flow).

This means investors still need to pay attention to the operating expense (Opex) guidance of tech giants, as efficiency improvements will be a key pillar supporting high valuations.

Debate Four: Agentic Commerce - The Ultimate Form of E-commerce

AI Agents will fundamentally change the way consumers shop. Morgan Stanley proposed the "5I" framework (Inventory, Infrastructure, Innovation, Incremental, Income Statement) to assess who can win.

Vertical Winners: Compared to general players like ChatGPT, those with specific vertical data and transaction loops (such as Amazon, Walmart, Instacart, DoorDash) will benefit first. This is because they have consumer trust and complete purchase histories.

Crisis for OTAs: Online Travel Agencies (such as Expedia, Booking) face significant risks. If Google or OpenAI launches an Agent capable of directly planning trips and making bookings, the traffic entry position of OTAs will be weakened, and the valuation system will be restructured.

Debate Five: The "Singularity" of Autonomous Driving Has Arrived

2026 will be a turning point for the availability of Autonomous Vehicles (AVs). The service coverage will leap from 15% of urban populations at the end of 2025 to 32%.

Fate of Uber and Lyft: The market worries that AVs will disrupt ride-hailing, but Morgan Stanley believes this concern is exaggerated. AVs will not kill Uber; instead, they will expand the entire mobility market by lowering the Cost Per Mile.

Key Indicators: Pay attention to the deployment of Waymo and Tesla in adverse weather (such as snow) and airport scenarios, as this is a sign of technological maturity.

Morgan Stanley believes that autonomous driving is not the end of ride-hailing but a booster for cost optimization. Uber remains Morgan Stanley's top choice

Debate Six: Embodied Intelligence - Amazon's Invisible Ace

While everyone is still discussing chatbots, the giants have begun to lay out "physical AI" - the combination of AI with robots and hardware.

Amazon's Robotic Warehouses: Amazon plans to add about 40 next-generation robotic warehouses by 2027. This is not just automation, but optimizing logistics paths and inventory management through AI. It is estimated that this could bring recurring cost savings of over $2 billion to $4 billion.

Unmanned Delivery: Uber and DoorDash are building ecosystems of drones and automated delivery robots, which will directly cut expensive labor delivery costs.

Morgan Stanley believes that investors should focus on companies that can use AI to solve the problems of "moving boxes" and "delivering food," as improvements in physical efficiency provide a stronger moat than virtual chat.

Debate Seven: The AI Revolution in Online Grocery

The U.S. offline grocery market is as large as $1.4 trillion, making it the largest potential gold mine for AI Agents.

High-frequency Necessities: Grocery shopping is extremely cumbersome and personalized, making it very suitable for AI Agents to intervene (for example, "Help me buy according to last week's list, but replace the milk with oat milk").

Winner Takes All: Amazon (Fresh), Instacart (CART), and DoorDash (DASH) are in advantageous positions. Especially Amazon, with its accelerated push in the fresh food sector and cost reforms, may become the next profit growth point.

Debate Eight: The Future of Search and ChatGPT's Advertising

Search is not dead; it is evolving.

Surge in Queries: The emergence of AI search engines has actually expanded the entire query market (TAM). Morgan Stanley predicts that the compound annual growth rate of search volume will reach 14% from 2023 to 2026.

Google's Moat: Surveys show that in commercial queries, Google's Gemini still has a slight advantage over ChatGPT.

ChatGPT's Advertising: As ChatGPT begins to introduce advertising, it will first impact those budgets that are not very effective and experimental (such as Pinterest, Snap), rather than Google's core search ads.

Debate Nine: The AI Revolution in Game Development

"World Models" are changing game development.

Cost Reduction and Efficiency Increase: AI-generated videos and interactive content will significantly lower the threshold and cost of game production.

Losers and Beneficiaries: This trend is long-term beneficial for giants with cloud computing power and AI tools (Amazon, Google), but for tool providers like Unity (U) and Roblox (RBLX), it is both an opportunity and a disruption risk.

It is worth noting that the AI transformation in the gaming industry is still in its early stages, and attention should be paid to those who can provide underlying computing power and generative tools, the "shovel sellers."

Debate Ten: The Redemption of Dating Apps

The online dating industry has experienced several years of stagnant user growth, and 2026 may be a turning point.

AI Matchmaker: Dating apps Tinder and Bumble are leveraging GenAI to improve matching algorithms and reduce users' "swiping fatigue."

Valuation Recovery: The valuations of Tinder's parent company Match Group and Bumble are already very low. If AI can successfully enhance user experience and restart growth, these two companies will see a significant valuation recovery