Cultivating AI Success

1. Cultivating AI Success

When I say, “AI first company,” John Deere is not the one that comes to most people's minds. They're a tractor manufacturer who primarily services farms. Its area of expertise is agriculture, one of the oldest industries on the planet. John Deere’s customers are intensely loyal.

2. John Deere

Why would this company get involved in AI? John Deere has two objectives. The first one is obvious. Farm equipment like tractors and combines can be automated. If the business can make those operate autonomously, it reduces the cost of farming. Given that there’s a significant labor shortage in farming today, AI can help offset the challenges farmers face with staffing their farms. The second objective is what’s called precision farming. Some farmers practicing this method see a 30% increase in crop yields. There's significant value and ROI from delivering AI to the farmer, one of our most traditional and oldest industries.

3. Where is the data coming from?

The first thing John Deere did was buy some startups. The business quickly ramped up its AI capabilities, but we need to ask the question that’s been the center of this section. Where is the data coming from? After identifying use cases, John Deere built the capabilities, but do they have the data? John Deere didn't, but they can access data-generating processes through their farming equipment.

4. Combining data

That's what the company leveraged to create data sets. John Deere introduced software into the combines decades ago, and this move towards automation was part of a much larger push. They were in the data-gathering phase, but the company’s approach led to conflict with their very loyal customer base. It was so contentious that customers did not want to buy new combines that included software.

5. Backlash

There was a significant backlash when it was revealed that some farm equipment and tractors had SIM cards to transfer data to John Deere. A secondary market was created by this intensely loyal customer base for equipment that did not include the new software and data-gathering features. Farming methods are considered trade secrets, and farmers are very protective of their methods. We must consider a new area when we start gathering data for analytics and models. Who owns the data? In this case, farmers generate data on tractors they bought and own. Who should own the data generated by those tractors?

6. Ethical considerations

John Deere delivered tractors to farmers who bought and paid for them. The farmer clearly owns the tractor. However, John Deere said it owned the tractor's software. By extension, it owned any data that the software gathers. John Deere contended that they granted farmers a conditional license to use the software, not ownership of it. The company laid out a clear argument from a legal standpoint, but is this data gathering ethical? There’s a more prominent theme at play here. Businesses operate with the myth of free data, but not all data is generated for free. There must be some collaboration and agreement in place. When data generators are employees, that agreement is present. When customers generate data, it’s a different story.

7. Who owns the data

Businesses must be aware of ethics and think about who owns the data. We're seeing more and more free and cheap data sources being cut off. Companies are charging more for them, like X and Reddit. Both social media companies increased the cost of access to their APIs because they learned LLM providers were training their models with cheap data from social media. When Apple cut off access to third-party data on its mobile platforms, Meta’s ad business took a hit. Data was John Deere’s ticket to fully autonomous vehicles and precision farming products. It began rolling both out in 2022 and has plans to keep releasing data and AI-supported products for the foreseeable future. As you plan access to data or data-generating processes, it's critical to consider the ethical implications of data gathering. Consider customer impacts. John Deere was fortunate not to lose its very loyal customer base. Some farmers say they bleed green because that's the color of John Deere tractors.

8. Future of AI within business

For other companies, ethical lapses will be a more significant threat. The ethical lapses will incur regulation, fines and penalties, but also loss of customers. As we chase the data and cultivate data-generating processes, we should be mindful of ethics and consider whether we truly have free access to data. Or if that free access could get taken away and threaten the competitive advantage or the future of AI within the business.