The AI Models That Are 100x Cheaper

Growth Newsletter: https://levelingup.beehiiv.com/subscribe

Need marketing help? Visit: https://www.singlegrain.com/ and https://npdigital.com/

Want to recruit great marketers? Find them here: https://marketingschool.io/hire

Neil spends 60 to 70% of his time reaching out to companies to buy, and explains when he wants founders to stay and when he does not, plus where global agency RFPs actually come from. Eric makes the case for open weights in marketing, where models like GLM 5.2 and Kimi K3 cut costs 25 to 100x on high-volume creative work. They then get into what nobody in his AI executive group could answer: how you actually measure the ROI of AI, why token usage is a terrible metric, and why revenue per employee is a better one. Closes on Jensen Huang's first post and Anthropic's MCP release finally being practical for non-technical marketers.

Key takeaways ◾ Open weights cut 25 to 100x on high-volume creative, and that is the marketing case ◾ Measure AI on revenue per employee, not on token spend ◾ The CEOs who spent last year cutting headcount have quietly backtracked

Chapters 00:00 How Neil splits his time: recruiting and buying companies 01:05 Buying businesses, and whether founders stay 03:17 Where the global RFPs actually come from 06:33 Why open weights matter for marketing 09:46 Measuring the ROI of AI 12:59 Sponsor break 17:00 Jensen Huang's first post and the open-weights letter 23:34 MCPs get practical