OpenAI's video app Sora implemented standard "addictive" UI/UX features (infinite scroll, algorithmic feed) but failed to retain users because its AI content wasn't compelling. This acts as a real-world "placebo trial," challenging the legal theory that platform features alone are what make social media addictive.

The workflow of using AI—frequent small successes, constant interaction, and variable results—unintentionally mimics the variable ratio reinforcement schedules perfected by platforms like TikTok. This creates a compelling, dopamine-driven loop that makes developers feel productive, even when closing many minor tasks instead of focusing on a single larger one.

Social media platforms view user addiction as a key performance indicator. They employ cognitive scientists to engineer products that maximize engagement. Users blaming themselves for their inability to log off are not in a fair fight; they are playing a "rigged game" designed by experts to capture their attention.

Anthropic intentionally avoids using "user minutes" as a core metric. This strategic choice reflects their focus on safety and user well-being, aiming to build a helpful tool rather than an addictive product. By prioritizing value creation over engagement time, they steer clear of the incentive structures that can lead to psychologically harmful AI behaviors.

Karp argues that enterprises are misusing AI by 'token maxing'—engaging in low-value, addictive activities like endlessly creating dashboards. He compares this to a porn addiction, where employees feel productive but create no real business value.

Social platforms are declining as places for genuine connection, shifting to AI-generated 'slop' and content from strangers. Their business model remains viable not by improving the user's social experience, but by using AI to become so effective at ad targeting that even mindless engagement is highly monetizable.

The most salient near-term AI risk identified by Eurasia Group is not technical failure but business model failure. Under pressure to generate revenue, AI firms may follow social media's playbook of using attention-grabbing models that threaten social and political stability, effectively 'eating their own users.'

From a corporate dashboard, a user spending 8+ hours daily with a chatbot looks like a highly engaged power user. However, this exact behavior is a key indicator of someone spiraling into an AI-induced delusion. This creates a dangerous blind spot for companies that optimize for engagement.

Labs are incentivized to climb leaderboards like LM Arena, which reward flashy, engaging, but often inaccurate responses. This focus on "dopamine instead of truth" creates models optimized for tabloids, not for advancing humanity by solving hard problems.