AI in Horticulture – Opportunities ahead for Business Management

2023 was the year generative artificial intelligence (GenAI) really broke through (check out the box below). It's wild how fast ChatGPT became a worldwide sensation and how it set off a chain reaction. Whether you're on board or not, it feels like we're on the verge of a new era where AI is shaking up the way we live and work.

People were saying similar things over 25 years ago when Deep Blue, a chess computer, beat world champion Kasparov. But today, we're also seeing an explosion of practical applications with real impact, even in horticulture.

When I started experimenting with ChatGPT and other easy-to-use tools, like those for generating images, I had this huge "wow" moment. The results were spot-on and surprisingly useful when you know how to ask the right questions. But the real game-changer is how generative AI integrates into your daily work—and Microsoft is leading the charge.

The potential of AI goes far beyond individual applications

Microsoft’s Approach

In recent years, Microsoft poured billions into a partnership with OpenAI, the company behind ChatGPT. With the launch of Copilot, Microsoft has made generative AI a core feature across its entire platform. Microsoft Copilot is like your personal digital assistant, helping with everything from writing documents in Word to creating dashboards in Power BI, or even developing apps in Power Apps.

Microsoft has invested billions in a partnership with OpenAI, the creators of ChatGPT. With the launch of Copilot, Microsoft has placed generative AI at the core of its platform. Microsoft Copilot functions as a digital assistant that helps users with tasks like writing documents in Word, creating dashboards in Power BI, or building apps in Power Apps.

The various Copilots are powered by the Azure OpenAI service, providing access to advanced language models, including ChatGPT. Unlike the public version, this allows businesses to train their own AI model using company-specific data, ensuring that data remains secure and protected, with enhanced security options available.

The potential of AI goes far beyond individual applications. Its most significant impact lies in core business functions. From inventory and sales management to finance and production, AI is set to bring greater efficiency and innovation to these areas.

Its most significant impact lies in core business functions. From inventory and sales management to finance and production, AI is set to bring greater efficiency and innovation to these areas.

‍

This video showcase sales order entry workflow for horticultural companies, utilizing Agriware 365 ERP and Microsoft Copilot (AI). It include instant order processing from emails, Copilot integration for drafting orders from various platforms, automated order sorting for accuracy, and a quick finalization and release process

Agriware Innovation Hackathon – Exploring AI for Horticulture

To explore the possibilities of generative AI for horticulture, we organized an innovation hackathon at Agriware. The hackathon focused on two key challenges:

• How can we develop new features that deliver tangible business value for our customers using generative AI?
• How can generative AI simplify our own work, such as programming, testing, documenting, and support?

The hackathon was a productive event that generated several promising prototypes in a short amount of time. Some of the highlights include:

• AWAI (Agriware AI): A smart assistant that optimizes production planning by analyzing Agriware data such as sales forecasts, availability, and production preferences.
• Sales Order Copilot: A tool designed to assist with creating sales orders more efficiently.
• Agrifolk Assistant: An interactive chatbot that improves user support in Agriware 365, enhancing overall usability

‍
Generative AI in Horticulture Operations

The hackathon illustrated the strong potential of generative AI for Agriware, and we anticipate a significant impact on the horticulture industry. The way we work is changing rapidly. Tools like Microsoft Copilot help users gain quick insights from specific contexts, making tasks such as data entry, report generation, and software development much more efficient and less time-consuming.

How to Get Started
To fully leverage tools like Copilot, preparation is key. This includes organizing data, training AI models tailored to your business, and ensuring that appropriate security measures are in place. It’s also essential to prepare your organization and employees for this shift in workflows.

While generative AI is powerful, it’s important to use it wisely. Starting with pilot projects allows you to explore its potential without taking unnecessary risks. The best approach is to "think big, start small, and scale up quickly."

If you’re interested in learning more, feel free to reach out to discuss tailored business management software solutions for horticultural companies.

‍About Mprise Agriware

Mprise Agriware provides custom, integrated Microsoft-based business management software for the horticulture industry. Combine finance, sales, inventory, production, and all your operations in one user-friendly app. Our software combines key business functions, helping to gain valuable insights and drive growth.

How Does Generative AI Work?

Generative AI creates new content, whether that’s text, images, audio, video, code, models, or test data. It uses different techniques, like Generative Adversarial Networks (GANs), where one neural network (the discriminator) evaluates the results of another (the generator) until it can’t tell the difference from the real data. Tools like ChatGPT use a method called Transformers, which has driven much of the progress in generative AI.

In the past few years, we’ve seen the rise of Foundation Models—AI models trained on tons of diverse data, making them versatile across different tasks. For text generation, these are called Large Language Models, like OpenAI's GPT-3.5 and Google’s BERT/Gemini. You can fine-tune these models with your own data for specific tasks, like those needed by your business.

There are a bunch of tools to interact with these models, including chatbots like ChatGPT and Google Bard, and other AI generators like Midjourney, DALL·E, Stable Diffusion (for images), GitHub Copilot, and OpenAI Codex (for code). The key to getting good results is the "prompt"—how you ask your question or give your command. So much so, that “prompt engineering” has become a skill in itself!

Pitfalls of Using Generative AI

As exciting as generative AI is, you’ve got to be careful, especially in business. One big issue is accuracy. Tools like ChatGPT can come up with highly convincing content that’s totally wrong (this is called hallucination). Plus, AI can easily generate fake images, videos, or mimic voices, which could be exploited by bad actors. So, blindly trusting AI outputs is risky—you always need to double-check the facts.

There’s also the issue of intellectual property and data privacy. While it’s super handy to generate summaries of sensitive documents or fix code errors, how do you make sure your competitors aren’t getting their hands on your proprietary information?

Other ethical concerns include biases in training data, which can lead to unfair results. Plus, these large AI models are energy-intensive to train and use, which raises sustainability questions.

So, while generative AI is an amazing tool, it’s important to stay aware of the risks and manage them responsibly!

This blog is written by dr. Cor Verdouw, Professor of Information Technology in Agri-business at Wageningen University. More information: https://research.wur.nl/en/persons/cor-verdouw

‍

‍