Friday, November 24, 2017

The future of independent crop consultants and killer ag apps

From:  Is Bayer’s Highly-Anticipated xarvio App the Next Uber for Ag?

"Yes, like I said this not only recognition, it's mixing four components in real-time in the nozzle. And we also are convinced that, speaking with regulatory bodies like U.S. EPA, we'll be able to register entirely new crop protection products which are resistance breaking, and we can tell them and actually prove that we are using them across only maybe 1% of the total acreage."

That seems really powerful and nothing short of just awesome. I've been talking for a long time about the convergence of big data and genomics and the impact on productivity and sustainability in this space and this is just another potential example of how this can all come together.  This also gets me thinking about an interview with Monsanto CEO Hugh Grant back in April 2016 where he discussed these kinds of applications in Monsanto's future:

"If I think about the next 30 or 40 years, I think through the use of data we’ll be applying these chemistries much more accurately and we’ll be applying them earlier, so applying them before diseases really take a hold in these crops or bugs are tearing these crops apart, so I think we’ll be more prophylactic, we’ll be more accurate and our selection of these chemistries will be a lot more discriminating. That’s kind of my vision of the future as through the use of data and bringing biology and science together, we’ll get much smarter about how we use these things, a bit like how the vision works for personalized medicine.” 

That sounds a lot like how they are describing  the Bayer xarvio app in the PrecisionAg article above. And, based on Grant's vision back in 2016, this all seems consistent with a strategy involving a Monsanto-Bayer merger.

However the article makes another interesting claim:

"The idea that a huge corporation like Bayer can largely automate and deliver agronomic recommendations remotely though a smartphone is a far greater danger to the future sustainability of any agronomic consulting operation, in this author's opinion, than a few large growers buying generic glyphosate off an online marketplace"

Based on what I have already said above, I can't doubt the value and potential disruption this app may bring with it. This makes me think of some parallels with apps and algorithms in the medical diagnosis space as well. Recently on LinkedIn data scientist Nasim Eftekhari offered an interesting perspective on this:

"With the recent news of the pneumonia detecting algorithm that has proved to be more accurate than human doctors, I see a lot of "who needs doctors anymore?" Or "are doctors going to lose their jobs?" Circulating around Linkedin. Guess who these algorithms learned from? Guess who labeled millions of images for the algorithm to find patterns? DOCTORS! What if new diseases and patterns emerge that are nowhere in the historical data? Who has to detect and interpret them? DOCTORS! 
AI can't replace doctors, but can tremendously empower them! It can give them a heads up when they don't notice a pattern they have seen before. Or for problems they have not seen before, it can help them with thousands of  similar examples and their diagnosis. It can give them access to the knowledge of smart doctors like themselves all around the world."

This reminded me of a discussion a few years back with economist Tyler Cowen on EconTalk with Russ Roberts:

"I would stress much more that humans can always complement robots. I'm not saying every human will be good at this. That's a big part of the problem. But a large number of humans will work very effectively with robots and become far more productive"

What Nasim says about doctors may largely apply to crop consulting, with some exceptions. Consulting might change, more focus on leveraging the array of tools, technology, and most importantly data being generated. As Grant noted back in the Here and Now interview:

"It takes about 40 decisions, from right around now until harvest in August or September, the grower takes about 40 decisions to produce a crop. Some of those decisions are highly technified, and others it’s because of what his mom and dad did or what he hears in the coffee shop or what he read in a magazine. So we’ve been populating those 40 decisions with data and I think by improving the quality of decisions, you increase the yield. I think the transition for Monsanto is increasingly in the next 10 years becoming a solutions-driven company, and coalescing the biology, the more accurate application of chemistry and the much smarter use of data. You know, these big green John Deere combines are streaming data off the field, one yard at a time, and it’s how you use that biological data and apply it back to the field to help growers with better insights, I think that’s going to be the next piece.” 

Taken one  decision at a time it might seem that a producer could cut out the middle man with a combination of a few apps. But they still might need someone that can step back and take a big picture systems view of the operation- that's where consultants can provide better insights and continue to create value. The way they do it just might not be the way I did 20 years ago with pen, paper, and a yardstick trotting across the field.

Sunday, November 5, 2017

CRISPR: Who’s pulling the wool over who’s eyes?

From: Amid GMO Strife Food Industry Vies for Public Trust in CRISPR (via NPR

(link)

http://n.pr/2yXk13G

"if this is genetic engineering, then call it that," says Perls. She says these producers are just trying to pull the wool over consumers' eyes with a strong public relations push."

Question: What do you call zapping a plant with radiation or dousing with chemical mutagens to create *random* gene edits vs. *targeted* ones? Well you can call it USDA Certified Organic, or slap a 'natural' label on it or have the butterfly logo assure you it's non-gmo. Who's really pulling the wool over who's eyes here?