ARK’s thesis is that consumer internet is moving from the web/mobile era into an agentic era, with AI shifting the operating system for shopping and search from query-to-answer to query-to-action. Nicholas Grus says 2025 marked the start of agentic commerce, where AI purchasing agents compress the shopping journey, could execute transactions in as little as 90 seconds, and rely on protocols like MCP, ACP, and UCP to connect retailers, agents, and consumers. He argues the biggest monetization pools are e-commerce and advertising, with AI agents rising from 2% in 2025 to 25% in 2030 and more than $8 trillion of online consumption flowing through them, while AI search grows from 10% today to 65% of global search traffic over five years. The closing speaker doubles down that this is happening faster than most believe and that human-to-computer interaction will look drastically different within a few years.
Nicholas Grus said the Big Ideas section is about transforming search, discovery, transactions, and e-commerce economics as AI enters a new agentic world.
Grus framed consumer computing as moving from the command era to the web era to the mobile era and now to the agentic era, with ChatGPT’s 2022 launch kicking it off.
He said the shift is from query to answer to query to action.
Grus said the two main consumer verticals for massive revenue disruption are e-commerce and advertising.
He said AI adoption is ahead of early internet adoption, with chat bots reaching almost 20% penetration in just over three years versus roughly seven years for the internet to get there.
Grus said 2025 was the year agentic commerce came to life, and AI purchasing agents are compressing the shopping journey from discovery to purchase.
He said a majority of shopping now happens before the actual purchase, and AI agents can personalize the journey, improve decision quality, and shorten transaction time.
Grus said the pre-internet shopping-mall or department-store purchase might take about an hour, while agentic-era transactions can happen in as little as 90 seconds.
He said agentic protocols are what make AI commerce real, because retailer back-end connections give agents the context and data they need.
Grus identified MCP as Anthropic’s model context protocol, ACP as OpenAI and Stripe’s agentic commerce protocol, and UCP as Google and Shopify’s universal commerce protocol, all aimed at enabling settlement and transactions.
He said the commerce stack has been unified from the retailer through the agentic layer to the consumer interface, whether that is a chatbot, voice assistant, or AI smart glasses, replacing the fragmented internet-era stack.
Grus said AI agents should capture a larger share of digital transactions as consumers delegate more commerce activity to them, with agents rising from about 2% in 2025 to 25% in 2030 and representing more than $8 trillion of online consumption globally in 2030.
On advertising, Grus said search is a roughly $350 billion market, Google dominates it, and Amazon has vertical-specific search positions.
He said AI can scale advertising by moving some search ads into the AI and agentic world, and AI search could grow from 10% today to about 65% of global search traffic over the next five years.
Grus said advertisers are slow to adopt new platforms, so monetization will lag and eyeballs do not automatically equal dollars; he also said OpenAI has started signaling advertising, making this year a first test of consumer response inside chatbot ecosystems.
He said current AI monetization is mostly subscriptions, with some enterprise consumption-based models and a bit of consumer consumption-based monetization, but the larger opportunity is indirect monetization via commerce take rates and advertising.
Grus pointed to social-media-like historical precedent that large-scale revenue mainly comes from indirect monetization rather than direct monetization, and he said the market could grow from around $20 billion today to around $900 billion in 2030.
He said AI lead generation ties advertising and commerce together, making those two areas the core consumer monetization opportunity inside what he sees as a new consumer operating system.
The closing speaker said the future is arriving faster than most believe, that human-to-computer interfacing will look drastically different within a few years, and that recent weeks have already shown that agentic change is happening now.