VCs say “super-app talk is mostly noise” as labs, agents, and vibe coders collide

The AI gold rush has a surprisingly boring truth: lots of companies are selling the same shovels. Labs pushing into coding platforms, coding platforms building agent-like apps, and startups trying to own the next workflow all amount to one thing. The supply chain is collapsing into direct competition.

One VC quoted by Business Insider summed up the mood: “Super app talk is mostly noise that's going to get resolved,” and the reason is practical. As foundation model players expand, they are not automatically winning the user experience battle. Instead, they are making bundles that may be “passable” at many tasks but not better than specialists, so users bounce back to what they already use.

Here is how the turf war is showing up in the real market. Companies that started with a narrower capability are widening fast. Last year, Anthropic and OpenAI launched Claude Code and Codex, AI coding platforms that rival Cursor and Cognition. And then things kept moving. Per user screenshots on X, Business Insider reports that Anthropic “may be working on an app builder for non-techies,” which would place it in direct competition with vibe-coding stars Lovable, Replit, and Emergent. Anthropic is a model lab, but the product surface is drifting toward app creation.

The overlap is not hypothetical. Business Insider notes that SoftBank and Lightspeed-backed Emergent said it saw Anthropic coming, and it even quotes Emergent CEO Mukund Jha. In an April interview with Business Insider, Jha said, “It's not a surprise. We've been anticipating this for a while and sort of internally thinking and preparing about it.” His bigger point is that coding is not the whole job. He said: “Coding is relatively like 20%-30% of the work. The hard work is actually taking the application to the last mile.” For non-technical users, Jha argues, secure and production-grade delivery is the tough problem that can get ignored when companies feel “spread thin” across too many initiatives.

While Anthropic is moving outward, OpenAI is making agentic moves. In February, the lab announced it was hiring Peter Steinberger, the creator of OpenClaw, an AI assistant builder that “skyrocketed in popularity earlier this year.” Business Insider frames this as a signal that OpenAI is joining a broader agent space that includes Sierra, a platform built by former Meta tech chief Bret Taylor, and Salesforce's AI agent platform Agentforce. The product evolution is also described directly: Codex has “evolved from a coding assistant to a virtual AI agent” that can parse and reply to emails, manage files, and schedule meetings.

This is the core “same shovels” dynamic. Everyone is chasing the same user value: execution. In the article, Business Insider gives more examples of overlap across categories. Emergent, which began as a vibe-coding platform, expanded into the personal agent space. Anthropic is also entering the design market, and graphic design giant Canva is pushing into a broader generative AI and productivity suite. These are not random expansions. They are attempts to capture revenue before others turn the feature set into a commodity.

That revenue pressure is where “super app” expectations collide with business reality. A vice president at early Lovable investor RTP Global, Tom Sheridan, tells Business Insider that the super-app fantasy is unlikely because the economics do not support infinite bundling. He ties it to the IPO clock, saying “Super app talk is mostly noise that's going to get resolved by the IPO calendar.” He explains the mechanism: foundation model players are in “P&L chicken,” and once companies go public, “cash burn stops being free” and “shipping into categories where you're good-but-not-best stops making sense.” In other words, the market may eventually force consolidation around winners.

This is not just about product strategy. Business Insider brings in a second risk that investors and board members should treat like wildfire. Dependency. Startups may build billion-dollar businesses on top of APIs controlled by companies that later compete with them. The article gives a concrete example: Cursor depends on Anthropic models to power features, but Cursor and Anthropic compete as coding assistant providers. That creates an awkward relationship where your core growth lever is also your future rival.

Even the data pipeline is at stake. Business Insider points out that big labs sprawling across directions can also shut out scrapers, affecting companies like Reddit or LinkedIn that host data, and harming smaller startups such as sales tech tools or meeting summarizers that need the data to improve. There is a silver lining for founders who understand what users want out of their data. Sheridan says foundation models can “see meeting transcripts shared for summarization,” but they do not know “which folders they should be filed under,” “what a team actually cares about,” or “the required follow-up actions.” That gap is where specialized startups can still win.

Finally, there is the consolidation question. Sheridan says: “I'd expect one of the major consumer AI breakouts to get acquired within the next 24 months, most likely by Google.” He adds that Google has the consumer ads business to absorb the cost and is structurally “the most desperate for consumer AI talent.” His conclusion is blunt about sequencing: “The first company to be bought gets the best price,” and “You don't want to be the last consumer AI play standing when each major buyer probably only takes one shot.”

For executives, the message is clear and a little uncomfortable. When labs, platforms, and agents keep moving up and sideways in the stack, differentiation shifts from “who has the model” to “who owns the workflow and keeps users in-session.” The shovel sellers will still find buyers, but only if they stop shipping bundles that users treat as temporary. Otherwise, the category goes from hype to churn, fast.