X Discussion: Jason Calacanis Says Under $40K a Year in AI Subscriptions Delivers $2M+ in Value and Has Automated at Least Seven Full-Time Jobs

Background

AI subscriptions and token spend are becoming a fixed cost for small teams, but there is little public, comparable data on how much real output they buy. At 05:27 UTC on November 2, angel investor and This Week in Startups host Jason Calacanis posted on X, opening with "THIS. IS. NOT. A. DRILL.", and shared his own math: under $40K a year in, more than $2M in value out.

What the post said

Calacanis wrote that he spends under $40K a year on tokens and subscriptions from Anthropic, Lovable, Grok Bot, ElevenLabs and a long tail of other products, and gets more than $2M in value at his investment firm LAUNCH and his show This Week in Startups. He said they have automated at least seven full-time jobs and become 100% faster at the average task, and that many things he spent hours on last year now take minutes. He added that the last 90 days changed more than the previous two years, and that anyone who has not used the last 90 days of products for at least 30 hours a week has no idea how profound this software is. A few hours earlier he had quoted Elon Musk's reply about Grok Bot's high serving costs, saying $5K a year can buy two or three competent mid-tier employees.

The discussion

At fetch time the post had about 2,201 likes, 106 reposts, 244 replies, 42 quotes and 230K views. Among supporters, entrepreneur Steven Cohn quoted it saying his team has seen the same level of ROI and that at 50 he has never seen anything like it; Farzad, with about 400K followers, quoted it saying "The ROI for AI is mind-boggling." Others were skeptical: one user mocked posting "we automated seven full-time jobs" with a siren emoji like a product launch and noted that 30 hours a week of usage is itself like a new full-time job, while others worried about their own jobs.

What it means for developers and teams

The figures are self-reported by Calacanis with no published method, so they are not an industry benchmark. But the post reflects a growing habit of budgeting AI spend like headcount. Teams that want a similar assessment can log time and cost per task before and after AI, compare that against subscription and token bills, and judge ROI from their own data rather than borrowing someone else's multiple.