by Dr. Saber Miresmailli
AI, AI, AI! Will Chat GPT take over all of our jobs?
What more can be said about AI? For the last several years it’s been the hot-button topic on everyone’s mind. Chat GPT is getting brought up by everyone from influencers, to business people, and even us farmers.
Is it good? Is it evil? What are its capabilities? What are its limits?
While these are all valid questions, I have a slightly different perspective on the potential of AI. Of course, this is revolutionary technology we’re dealing with, but I do not think it is something that should be feared, or revered. At least not in every context and domain.
Instead, AI should be seen for what it is – a tool that can be used in tandem with other technologies to provide value. While AI has had a lot of buzz surrounding it recently, I don’t think we should see it as any more important than the countless other pieces of technology that make up our world.
Keep reading for a more in-depth view of how I view AI, how ecoation uses this technology, and whether or not we should be considered an AI-based company.
How I view AI
The term AI has become abused lately. Let me explain.
Any company that has statistics, computation, or machine learning involved in what they sell has been quick to label itself as an AI company. This is what happens when a certain aspect of technology becomes overwhelmingly trendy. Even if a company is using a combination of technologies that don’t involve AI, by labelling them as such they can get more attention from the public, and from investors.
It’s easy to begin labelling every piece of technology as AI because it will grab people’s attention. There is not nearly as much buzz surrounding other technology, like machine learning for example.
One interesting thing about AI is that like many new and exciting technologies, it is like a double-edged sword. To the general public, AI comes with just as many negative connotations as it does positives.
AI can learn your patterns and behaviours. It can mimic people’s voices and appearances. Take a look at recent news stories of kidnapping scams using cloned voices to become more believable. That’s scary. Then, think of this technology's potential to sway the mind of a populace or impact global politics. That’s terrifying.
In this sense, the same way that calling your technology “AI” can get you some brownie points with investors, it may ultimately turn them off instead. Some end users will not want to use a product that labels itself as AI for fear that it will spy on them.
This becomes quite funny when you realize that many of the companies labelling themselves as AI are using technology that is no more sophisticated than the water adjustment function of a washing machine. Are people worried that their washing machine is going to spy on them? Of course not. But if you labelled it as AI, people would begin to have concerns.
This is the power of public perception of technology. This is also why the accurate distinction between different technologies is so important.
How ecoation uses AI
When it comes to ecoation, we are a company that uses classic hardcore statistics, machine learning algorithms, as well as AI to make our service function the way it does.
The purpose of AI in our company is to extract patterns, learn from these patterns, and create a level of understanding that didn’t exist before. It’s worth noting that this has been a technology long before the term AI became sexy. We used to call them decision support systems, or DSS. AI just take this a step further and in some occasions, automate the decision based on the data.
While we do use AI as a function of our company, I don’t believe we are exploring the scary side of the technology. Instead, I believe we are exploring the exciting side.
The difference between regular old-school statistics and more advanced AI systems is that AI systems can consume massive amounts of data in a short period of time. As this occurs, the AI can begin to associate events with their results. This only becomes more accurate the more data is provided.
It’s crucial here to not let our datasets become biassed, as that will alter the interpretations the AI can make. To do this, we need to understand where the shortcomings are in these AI models. In reality, it’s not too difficult to sway an AI model’s outcome if you are inclined to do so.
Number one, we try to make sure our dataset is as diverse as possible. We have trillions of data points collected, but it is not just about the numbers. What makes all these data points so useful is that they come from all over the world – Japan, Mexico, Italy, Canada, everywhere.
Let’s say you were to collect vast amounts of data from Canada alone. The conclusions you draw from that data would be biassed towards Canada, and may not reflect results from other countries. However, by giving an AI model, data from all over the world, it will better be able to interpret the differences caused by geographic location.
At ecoation, we have a standard data collection solution. By building our hardware in house, we can be certain that the data we’re collecting is standardized across all greenhouses. We can be confident that the differences in data are due solely to geographic diversity, and not fluctuations in measurement accuracy.
By having a dataset that is as unbiased as possible, we can be confident that our AI is making correct assumptions about the information we provide it with. What is noteworthy and important here is that we dont combine data from multiple sites and we dont share customers data. We use the diverse data set that we have to enhance our learning and algorithm performance.
We have two systems in place to ensure that our datasets are consistently accurate. First are the automatic checks and balances built into our algorithms to ensure data is being collected properly. Second, and perhaps even more important, is our excellent customer success team which is in constant communication with each of our clients. By having weekly conversations, they can determine when there is an anomaly, and take the necessary actions to fix it.
Essentially, by having a massive, accurate dataset, our AI can draw conclusions that are as accurate as possible. It may not be the sexiest use of the technology out there, but it provides value, which is always our priority.
Why I do not see ecoation as a 100% AI-based Company
This brings us to the question: is ecoation a 100% AI-based company?
I am very hesitant to label our company as being 100% AI-based despite our use of the technology. While this is partially due to the potentially negative conceptions of the technology, the real reason is quite different.
The reason I don’t like saying ecoation is a 100% AI company is because it is so much more than that. While AI provides a lot of value to our business, it is just one piece of the puzzle.
While many companies are willing to label whatever technology they use as AI because it sounds good, I believe this is a disservice to the value that other technologies provide. Our use of AI to analyze data wouldn’t be possible without the use of statistics, computation, machine learning, the engineering behind our hardware, the people who make it run, and so much more.
By calling ourselves a 100% AI company, I believe we risk losing sight of all the things that make ecoation special, beyond just one aspect of our business.
With many innovative technologies – AI just being the latest example – people like to view them as either black or white. This ignores the many shades of grey in between. AI is not a malevolent force that will destroy us, and it also shouldn’t be seen as the only technology that matters. There often isn’t much reason to get too excited, or too scared about any given technology.
In reality, it’s somewhere in the middle. At the end of the day, AI is just another tool in the arsenal for companies like ours to provide value to our customers. While I do think there are both exciting and scary applications for AI, I don’t want to let it distract from everything else that goes into making our company run.
By simply calling ecoation a 100% AI company, we are doing ourselves a disservice. This is because the work we do in other areas – from our mechanical design to our user interface, and so much more – is amazing. I don’t want to let one trendy aspect of our business distract from this.
With both ecoation and any other companies that use AI, I challenge you to look a bit deeper. Try and see beyond that single layer, and appreciate all the different technologies at play that often get ignored in favour of something more exciting.
This will not only give you a more accurate view of what makes companies like ours function, but a deeper appreciation for all the technology that makes up our daily lives.
Buttom line is that not every company is Chat GPT and Cambridge Analytica. While we might share some tools and parts, we are not completely the same. You might find a M8 screw in your car and in a fighter jet! It does not mean that you can break the sound barrier with your hot wheel — although knowing my good people in the Greenhouse industry, I bet a few of you have tried!