Let's look under the hood of the technology powering AI
Today “everyone” is talking about AI and there are many words that are used more or less correct in the public discussion on AI. In today’s Agtech Letter we will start untangling all these concepts and words to improve your insights into what AI is.
First, lets talk about some of the words
During the last 5-10 years artificial intelligence, or AI in brief, have been used as a buzz word in many industries including academia. The use of buzz words are good in the sense that many can relate to them and can associate the words with some property or behavior. However, the bad thing about buzz words are that everyone can have their own understanding of what it is and there is no clear definition of the buzz word.
An example related to AI is the difference between artificial general intelligence (AGI) and narrow artificial intelligence, where AGI is the type of AI that people refer to when discussing how computers can be as smart as humans and have capabilities to solve a wide range of problems. On the other hand, narrow AI is a solution with AI technology that is used to solve a specific problem that it has been trained to solve.
Most researchers agree on the fact that we only have narrow AI today, and no AGI solutions is developed yet at the time of this writing.
Common words that are used for AI technology
As already mentioned, lots of different words are used when talking about AI, and to follow the discussions and descriptions of new solutions it is good to have heard some of these words to relate them to each other and understand roughly what is meant by them. Artificial intelligence is a wide area with lots of technologies and applicartion areas, but one of the most common concepts in AI is machine learning, which is a type of algorithms that can be used to “learn” behaviors from data.
Machine learning algorithms can be of many different types, and what type that is used is mostly dependent on the type of application it should be used in and what data that is available. However, a common type of machine learning algorithm that has become very popular and that many have heard of are neural networks that try to mimic how our human brains work.
If several neural networks are combined, you get something called a deep neural network, and algorithms that use deep neural networks are called deep learning algorithms. Deep learning can be used in many different application areas such as finding weeds in pictures if crops, determine if there is an obstacle in front of a tractor, and chat bots such as ChatGPT. Each application area of course require its unique tweaking of the algorithms, but overall it is based on the same technology.
So what is the value of knowing all these words if you work with crop production or dairy production? Well, to me its all about understanding on a high-level what the technology is all about and how it applies to you and your farm in order to make smart decision for earning money on the opportunities while keeping the risks low. And this understanding starts with understadning some of the main words and their relation.
Agtechers' Actions
Many often use the term AI very wide and loosely defined, and that’s OK. But when you evaluate and scout for new AI solutions, be sure to ask the right questions on what type of technology that is used and how you can earn money by investing in the technology.
In the next Agtech Letter we will provide part two of the technology that powers AI.
