Jul 15, 2017 · 15m · a16z
Chandra Krintz
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
At the Andreessen Horowitz Academic Roundtable, Professor Chandra Krintz presents UCSB's SmartFarm initiative, which adapts consumer data analytics and edge computing systems to deliver scalable, autonomous precision agriculture tools for modern growers.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 0.0 Guest teaching 0.0 Guest disagreement 0.0 The host pushing back 0.0
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 13:50 Critique of corporate data harvesting
Krintz firmly criticizes tech companies for taking farmers' valuable data for free, insisting growers must retain full choice and control.
Hardest push from the host ▶ 0:18 No host pushback (monologue format)The episode is a solo presentation where the host does not speak, resulting in no host pushback.
Biggest teaching moment ▶ 9:50 Explaining farm connectivity realitiesKrintz educates the audience on the severe internet bandwidth limitations on farms that necessitate localized hybrid computing.
The host holds their own ▶ 0:18 No host hits back (monologue format)Because this transcript consists entirely of a guest presentation without host dialogue, there are no host expertise counter-points.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| Analytics in the Modern Economy Versus Agriculture | 0 | 0 | 0 | 0 | This is a solo presentation monologue where Chandra Krintz discusses consumer analytics and the urgent need to apply big data solutions to agricultural production. Because the host does not participate, all host-side scores and combativeness remain at zero. | |
| Adapting Consumer Analytics into Agricultural Implements | 0 | 0 | 0 | 0 | Krintz outlines practical applications of analytics to agriculture, referencing interpolation algorithms and joking about creating Moogle for livestock. The host does not speak in this segment. | |
| Aerial Imaging and Irrigation Thermal Leak Detection | 0 | 0 | 0 | 0 | The speaker explains aerial thermal imaging for detecting irrigation leaks and introduces four key computer system research areas. Host scores remain zero for this monologue section. | |
| Hybrid Cloud Architecture and Farm Connectivity Challenges | 0 | 0 | 0 | 0 | Krintz discusses farm internet connectivity constraints and introduces a low-cost, portable hardware appliance for on-farm edge processing. An audience member briefly interjects agreement, but no host interaction occurs. | |
| Repurposing Cloud Fault Tolerance for Edge Computing | 0 | 0 | 0 | 0 | The presentation details shrinking cloud architecture for fault-tolerant edge computing and emphasizes data ownership rights for growers. The segment is entirely monologue. | |
| Acknowledgments, Community Call to Action, and Conclusion | 0 | 0 | 0 | 0 | Krintz thanks her team and sponsors before closing with a call to action to build an open community to help feed the world. Host scores remain zero. |
Statements from this episode (14)
Agriculture is completely underserved by modern technology and analytics
“If you think about the ag industry today, it is completely underserved by technology, let alone analytics.”
The global supply of arable land is fully exhausted
“There's, the amount of arable land in the world is done. We're, we've used it all.”
California produced $50 billion in agricultural output last year
“The amount of food we produce in California alone is enormous. Fifty billion last year.”
Precision agriculture needs Amazon, Google, and Tinder-style analytics
“So in terms of analytics then, we need to make Amazon recommendations to farmers for things like irrigation scheduling or the application of inputs, fertilizers, pesticides, water. We need Google search for crop disease. We need Tinder for pests.”
SmartFarm identified crop quality root causes via diffusion analysis
“We took the same data as part of our Smart Farm effort, and we did a diffusion analysis to try to understand not how to control a tractor, but why this differentiation of quality was happening. So we applied this, a diffusion analysis technique to figure out t…”
Spatial analytics allow farmers to target problem areas efficiently
“Once you have this, something like this, you can hand a farmer an iPad, and they can, you know, instead of walking the entire field, they can focus their energy to find where the problem might be.”
Thermal imaging detects agricultural irrigation leaks via temperature variations
“You can also use this type of imaging in terms of for using the thermal spectrum to identify where leaks occur. Because leaks have an area that is a little bit different temperature than the rest of the field.”
On-farm edge computing must require zero system administration
“If we're going to put some computation on farm, the system that we develop there must be lights out, easy to use. It must be self-managing, must be completely fault tolerant, must be like a TiVo box. You turn it on, it just works. Right? And you don't need a s…”
Most farms lack reliable, low-cost internet connectivity
“Most of our farms are lucky if they are connected to the internet. And if they are, it's extremely low bandwidth, and it's very expensive.”
SmartFarm edge appliance runs AWS and Google Cloud workloads locally
“Anything that runs in Amazon and anything that runs in the Google Cloud platform will run on this appliance.”
Analytics frameworks like Spark and Hadoop assume exclusive resource access
“Most things, Hadoop, Spark, Storm, they think they're running by themselves. And so they compete for resources in really interesting ways.”