Danny Bernstein left Big Tech to fund farm-bots | E2310

Episode Summary

Executive Summary: The episode centers on agtech startup Reservoir Farms and founder Danny Bernstein’s effort to build a physical incubator where startups can test robotics, automation, and chemical-reduction tools on real farms. It then pivots to a debate about AI’s impact on design and education, arguing that AI raises the baseline while increasing the value of great human work, but also exposes fraud, cheating, and misaligned incentives in schools.

Main Topics: Reservoir Farms and the agtech incubator model: Danny Bernstein explains how Reservoir combines an operating farm, incubator, and venture fund to help agtech startups build and test products on real land near Silicon Valley. Labor shortages and automation in specialty crops: The discussion emphasizes that specialty crop agriculture is highly labor-intensive, domestic labor supply is insufficient, and automation is needed for hard-to-fill jobs like strawberry and grape harvests. Chemical reduction and biological alternatives in farming: Reservoir’s focus includes reducing chemical use and replacing pesticides with technologies such as electric weeders, laser weeders, steam fumigators, UV light, and autonomous precision application. Business model and economics of the farm incubator: Bernstein details the dual structure of Reservoir Farms and Reservoir VC, including rent-paying startups, operating revenue, and a small proof-of-concept fund designed for sustainability. AI, design, and the Figma debate: Jason and the host revisit a Twitter dispute about whether AI threatens design, concluding that AI raises the average quality of design work while making strong design even more valuable. Higher education accountability and student debt: The show critiques college programs that leave graduates unable to repay loans and argues for stronger accountability on schools, especially in low-ROI graduate programs. AI-assisted cheating and the limits of remote assessment: A professor’s data showing large score gaps between AI-assisted take-home midterms and in-person finals is used to argue for in-person testing and better human oversight.

Key Arguments: Specialty crop farming is a labor crisis: domestic workers largely avoid physically punishing field jobs, so automation is necessary. Agtech has historically scaled slowly because startups lacked easy access to real farms; Reservoir fixes that by giving them immediate physical access. A sustainable agtech incubator must be run like a real business, with recurring revenue and clear operational costs, not just venture hype. AI does not eliminate design; it raises the baseline quality, making genuinely excellent design more valuable and increasing the moat for strong platforms like Figma. College and graduate programs should be judged by economic outcomes and loan repayment ability, because taxpayers are effectively backing the debt. The current education system and many assessments are broken because students can use AI to cheat and teachers often fail to enforce rigor. AI tools should be used as supplements that improve human output, not as replacements for thinking, writing, and analysis.

Data Points: Farm jobs posted last year: over 400,000 - Referenced to show the scale of agricultural labor demand. Domestic applicant rate for farm jobs: less than 1% - Used to argue that domestic workers generally do not want these jobs. Labor share of specialty crop production costs: 40% - Bernstein says labor still makes up a large portion of costs in specialty crop agriculture. Labor share in strawberries: 60-70% - Illustrates how labor-intensive strawberry production is. Labor share in table grapes: 80% - Used to show extreme dependence on labor in some crops. Time from raising VC to farm access: 9 months - Reservoir surveyed startups and found it often took nine months to get onto a real farm. Reservoir VC proof-of-concept fund: $3 million - The fund wrote five checks and is being re-upped. Reservoir Farms ARR forecast: $5-7 million - Bernstein says the operating company expects this annual recurring revenue range. Full-time team size: 10 people - Reservoir Farms has a small operational team. Startups paying rent: 20 startups - Companies pay to use the farm and facilities. Monthly rent per startup: $3,000-$6,000 - The incubator charges startups for physical access and support. Farm footprint: 40 acres - Reservoir Farms’ main Salinas campus. Total acreage operated: over 100 acres - Bernstein notes the company now operates multiple farms across the American West. Programs that may fail new education test: 800,000 students - Department of Education estimate for programs that would not show positive earnings outcomes. Undergraduate certificate programs that would fail: 18% - Cited in the education accountability discussion. Figma valuation: $11 billion - Jason references this when explaining why he bought shares. Figma trading multiple: 7x sales; about 70x P/E - Used to justify the investment thesis. AI-cheating delta example: Student 52: 67.5 midterm, 100 final - Illustrates the gap between AI-assisted and in-person performance.

Pivotal Quotes: "There are over 400,000 farm jobs that were posted last year, and less than 1% of them received a domestic applicant." — Host: Opening argument framing the labor shortage in agriculture. "I call it the white whale of ag tech." — Danny Bernstein: Referring to strawberry harvest automation as a long-sought but difficult problem. "If a program cannot show that it leaves its graduates financially better off than if they had never enrolled, it should not be underwritten by federal taxpayers." — Nicholas Kent (quoted in discussion): Summarizing the education accountability rule.

Implications: The episode suggests agtech will accelerate where companies can test on real land, AI will commoditize average design while rewarding top talent, and universities will face rising pressure to prove ROI. It also argues for more rigorous in-person assessment as AI adoption grows.