10 Insights for Agribusiness Leaders from Mary Meeker’s Artificial Intelligence Report

The 2025 Trends – Artificial Intelligence report is authored by renowned analyst Mary Meeker and BOND Capital.

The report is a comprehensive overview (340 slides) of the current state and trajectory of AI adoption globally. The report captures the velocity of change, the evolving competitive landscape, and the technological changes that are impacting how work, decision-making, and infrastructure investment are being redefined across essentially every industry.

Mary Meeker is a legendary technology investor and analyst. Her annual Internet Trends reports became essential reading for executives, investors, and policymakers for being rigorous and including exceptional foresight. Now, she is evolving from internet utilization to artificial intelligence and the learnings are fascinating.

I wanted to take the most relevant insights from the report and apply them directly to agriculture, with a focus on what they mean for agribusiness professionals and companies across crop input manufacturing, ag retail, equipment, and agtech startups.

Below you’ll find 10 key insights from the report and the considerations and implications for agribusiness professionals.

Index

1. AI Adoption Is Outpacing Historical Tech Trends

2. Falling Inference Costs Enable More Use

3. Multimodal AI Is Growing and so is Native Integration

4. AI in the Physical World Is Moving Fast. Autonomy Is Coming.

5. Enterprise Adoption is Growing and Is Revenue-Driven, Not Cost Management Driven

6. AI Model Performance Now Exceeds Human Levels in Key Domains

7. Developer Ecosystems Are Exploding

8. AI Reduces R&D Cycles

9. AI Will Not Be Optional Infrastructure. It Will Be Mandatory.

10. Carbon Robotics Acre Treatment Highlighted

11. Things I Still Wonder

1. AI Adoption Is Outpacing Historical Tech Trends

ChatGPT reached 800M users in 17 months, faster global penetration than every major technology, ever, which illustrates how fast adoption can move when the infrastructure (devices, internet, and apps) and instal base is already built out.

The CapEx spend by the Big Six tech companies hit $212B in 2024— up 63% YoY, mostly to support AI infrastructure:

ChatGPT is doing 365B+ annual searches after just 2 years, 5.5x faster than Google’s historical growth.

Implication: Agribusiness professionals must assume that AI expectations (from farmers, distributors and internal teams) will ramp faster than past tech waves. Consumer expectation will be set and agribusiness will need to follow that lead. Additionally, for marketers, the old “SEO” logic and information flow needs to change, as my friends at Pay Dirt Digital have talked about.

2. Falling Inference Costs Enable More Use

While AI training costs are skyrocketing, inference costs per token are plummeting, by upwards of 99% from 2020 to beginning of 2025. Inference is the process of using a trained AI model to generate outputs or make decisions based on data.

As an aside, an interesting takeaways is that AI Usage is still growing faster than revenue. OpenAI revenue is rising ($2–4B est. in 2024) but compute costs still outstrip revenue, burning billions.

Implication: On-device or edge-AI for agronomic decisions (eg: disease identification, variable rate prescriptions) will become cheaper, faster and more widely accessible. An example of utility is Sentera and their SMARTSCRIPT Weeds offering, now apart of John Deere, and enhanced turn around and in-field capabilities.

3. Multimodal AI Is Growing and so is Native Integration

Models like GPT-4o now process text, image, voice, and video inputs natively, with growing realism in conversations, image generation, and even voice cloning.

Implication: User interfaces and digital workflows will need to support non-text AI interactions. This means companies must prepare for a shift from GUI-based products to voice, and even vision-native experiences. In a recent edition of Upstream Ag Professional, I went through The Promise of GenAI: Invisible Technology That Fits How Agriculture Already Works and talked about the utility of embedded GenAi.

4. AI in the Physical World Is Moving Fast. Autonomy Is Coming.

Subscribe to Upstream Ag Professional to read the rest.