Cheating at Cheating? Cluely's $120M Bet on Proactive AI

Cloy’s $120M bet isn’t really about “cheating”—it’s about building the next wave of proactive, agent-style AI that quietly sits across everyday apps and becomes a daily habit. The controversy around “cheat” branding may drive attention, but the core value being highlighted is the product experience: level two proactive AI agents delivered through a UX that feels more like an always-on assistant than a chatbot.

The transcript draws a sharp distinction between marketing and model quality. Cloy’s AI intelligence is described as “mid,” with the experience likened to “ChatGPT-4.0 level intelligence.” Still, the practical usefulness comes from how the system behaves: it can help users move forward, including tasks like watching and discussing a YouTube technical topic, troubleshooting web design issues (such as CSS), and generating suggested code based on what it can observe on the user’s screen. The emphasis is that the agent’s proactive workflow—its ability to act in context—matters more than raw model brilliance.

That framing connects to why the “cheating” narrative is treated as a deliberate distribution strategy. The company’s manifesto (“cheat at everything”) and the repeated press cycle around founder Roy Lee’s alleged cheating-related background are portrayed as high-impact, emotionally resonant messaging aimed at Gen Z and Gen Alpha. The transcript claims the team is intentionally using aggressive marketing—seven marketers versus four engineers—to win mindshare and earn clicks, with the goal of making Cloy feel invisible but indispensable, like a pane of glass over other apps. The long-term bet: once users form daily habits, they won’t easily switch, even as newer models arrive (it name-checks future “ChatGPT-5” and “ChatGPT-6”).

A16Z’s involvement is positioned as validation of the product and distribution plan rather than the controversy. The transcript suggests the investment targets both consumer habit formation and enterprise adoption: Cloy is said to be cutting “seven-figure deals,” supporting customer success and sales workflows by letting teams share screens while Cloy runs in the background. In this view, the “cheating” label is a hook; the underlying product is an agentic layer that improves workflows without forcing users to treat AI as a one-shot answer machine.

Finally, the transcript argues that the real competitive shift won’t be about whether AI can cheat—it will be about proactive agent design and UX that can stay ahead of model capability. As proactive tools become widespread, the “cheat code” advantage fades, similar to how resume-writing became commoditized after ChatGPT adoption. What remains valuable is the ability to open the distribution window early, deliver proactive experiences, and potentially unseat chatbot-first workflows. Cloy is presented as an early, instructive example of that proactive AI revolution, even if its branding remains polarizing.