Let’s pause from the daily toils of ag for a minute and start thinking about the future. In a very short time artificial intelligence (A.I.) is coming into the precision ag space and we must ask ourselves how we are going to utilize it. Now we must overcome our initial fear and skepticism of the world turning into some sort of sci-fi horror story that is straight out of The Matrix or The Terminator movie franchises and start to see the ways that A.I. will radically change our lives for the better. We’ll take a look at what the future might look like in a bit, but first let’s go over a few definitions and concepts.
Machine Learning vs. A.I.
If you are like me, you keep hearing these buzzwords like “A.I.” and “machine learning” and keep wondering to yourself “what the heck are these and what is the difference?” It can be hard to keep up with all of this jargon, especially with the worlds of precision ag and computer science merging together at an ever increasingly rapid pace. It seems like these two terms have been used interchangeably, which has led to some confusion. Here are the best definitions I have found for these two terms:
Artificial Intelligence: The development of computers to do things normally done and associated by people who are acting intelligently. Some examples of these sorts of tasks are problem solving, reasoning, and perception.
Machine Learning: A subset of the broader artificial intelligence space where machines have access to data and are able to analyze and learn from the data available (which could be a lot). Machine learning came about because engineers realized that rather than writing enormous files of code to program computers and machines on how to do everything, it would be far more efficient to code them to think like human beings, and then plug them into the Internet to give them access to all of the information in the world.
Now that we have a basic understanding of these concepts, let’s talk about the different types of A.I., as well as the different A.I. programs that are out there today.
Weak A.I., Strong A.I., and Watson
There are currently three groups that A.I. systems fall into. The first is referred to as “weak A.I.” that are aimed at simply getting systems to work and behave like humans, but ultimately don’t think like humans. An example of this is IBM’s Deep Blue, which is a system that was a master chess player but did not play the game in the same way that humans do.
The second simulates human reasoning and is referred to as “strong A.I.” In theory, strong A.I. systems are designed to think like humans do. I say “in theory” because a true strong A.I. system has yet to be developed and there are some doubts if it is even possible to be able to simulate human cognition, complete with our complex emotions, by a machine.
The third is a mixture of these two systems. These systems use human reasoning as a sort of guide but are not trying to model it perfectly. This is where some of the biggest advances are being made today and this group is where IBM’s Watson falls into. Watson works by scanning thousands of pieces of text for patterns and then adds up all of the evidence that ultimately give it a conclusion that it is confident in. An example of this is the application of Watson into H&R Block’s tax preparation program. Watson more or less reads the enormous tax code of the U.S. then applies the proper information to your tax return to maximize your refund. This is great because Watson is maximizing your return by doing something efficiently and effectively that even the best accountant couldn’t realistically imitate because of the sheer volume and complexity of all of the tax codes.
So What Does This Mean For Precision Ag?
As we get deeper into and more dependent on computer programs to deliver insight about our business/farming operations, it seems only natural to layer the pieces of data together. Once we do that, we can learn some deep insightful things that we may have never realized before. And as we dig deeper, we recognize that the amount of data out there to analyze is more than the ocean is deep. With that being said, if Watson can handle complicated tax returns effectively, think about what type of analysis that could bring to the multiple layers of data collected on your farm.
I think once this technology is brought to the farm gate we are going to quickly realize how inaccurate a lot of our data layers are from our yield data in many cases down to our soil data, but that’s a different topic for a different day.
Looking towards the future, I imagine being able to detect insect and disease pressure from aerial imagery and then overlay that scouting information with an advanced weather model to not only predict the movement of pests, but also to alert areas that need to take preventative action. Using all of these layers of data together will ultimately maximize their value and that will be difficult to do correctly without the help of artificial intelligence. If that comes to fruition, would this change how we market grain? What would happen to volatility in the market? Would this change profitability? These are very interesting questions and there is no doubt that A.I. has the potential to be a major game changer.
Will A.I. Eventually Take Over the World?
I think we’re too focused on Hollywood theatrics and conspiracy theories about the HAL 9000 artificial intelligence system from 2001: A Space Odyssey that turned homicidal after it was discovered that it had made an error. Some conspiracy theorists note that each letter in HAL is one letter away from IBM. A.I. is not sinister at all but rather it is proposed to become a valuable tool at our disposal to help propel the human race into the future in hyper speed, so let’s all calm down with this conspiracy hogwash.
A bigger concern of yours is that A.I. might very well replace your job somewhere down the line. Apple just recently got approval to test their self-driving cars on California roads. I think the writing is on the wall for much of the transportation industry and that could trickle into agriculture before long. We all saw the tractor without a cab at the Farm Progress show last year, but that seems like a stretch for the foreseeable future since it not only would be hard to get that vehicle from your farm to the field but just try to find someone who doesn’t enjoy driving a tractor and doing field work.
But will we ever see a fully autonomous aerial applicator? Surely that would make sense eventually given the danger of that profession, assuming we can get past all of the bureaucratic red tape in the aviation industry. What about Watson replacing the role of the crop consultant? Now that has the potential to be a real market disruptor!
Time will tell on all of these, but there is no doubt that times change and so do jobs. When was the last time you saw a blacksmith? Sure there are a few out there, but it’s a far cry from the number there were 100 years ago when draft horses were essential farm equipment. The moral of the story is that change is coming and you can either embrace A.I. or get run over by it. The choice is yours.