[ad_1]
Mike is a farmer in his 40s from South Queensland. With a chestnut tan, an overwhelming handshake and a strong Outback accent, he is the third generation in his family to grow sorghum, a grain primarily used for animal feed.
But, like most farmers, Mike faces more challenges than his ancestors. Climate change has eroded the profitability of Australian farms by an average of 23% over the past 20 years. Improving productivity by producing more with less is a constant challenge.
Read more: Australian farmers are adapting well to climate change, but there is work to be done
After the devastating bushfire season of 2019, Mike began exploring “smart†farming techniques made possible by artificial intelligence (AI). Agriculture is considered one of the most fertile industries for AI and machine learning. Mike was excited about an AI-powered system that allowed him to use less fertilizer and water.
After months of investigation, he found a company promising that its technology could reduce farm inputs by up to 80%. This involved software processing of information from digital sensors placed in his fields to enable “precision farming†– tailoring the treatment of water, pests and fertilizers to each plant.
The sales pitch was convincing. But the cost of installing the system was $ 500,000, plus $ 80,000 per year for data storage and processing. Assistance costs were added to this.
In the end, Mike calculated that the cost would outweigh any additional profit generated, even if the nifty technology kept all its promises. If he delivered less, that would only help him go bankrupt.
This experiment – of being touted as AI technology with big claims but questionable value – is commonplace. It’s easy to get swayed by promises. But new technologies are not the solution for everything. To make it worth it for people like Mike – indeed any organization – requires a cold calculation of its economic value.
In this article, we offer a simple methodology to do so.
Blinded by technological potential
Despite all the attention paid to how AI will revolutionize the world, the hype about it is nothing new. Since the inception of practical AI techniques in the early 1960s, obsession with the potential of AI has led to two major “AI winters” – in which the huge investments of companies and research institutes did not deliver the results promised.
The first took place in the 1970s, when money was pumped into various AI systems such as speech recognition and machine translation. The second came in the 1980s, when companies invested heavily in so-called “expert systems” to do things like diagnose diseases or control space shuttle launches.
PA
Either way, what the technology could do was well below the hype. It wasn’t that AI was useless. Far from there. But what he could do had limited economic value.
The backlash set back the scientific and economic advancement of the technology by nearly a decade each time, as funding and interest waned.
To make sure that your investment in technology is worth it, you must be careful not to get carried away by the promise and the possibilities.
As Ben Robinson, chief strategy officer at financial software firm Temenos, said:
we can safely predict that it won’t be blockchain, APIs, or AI that will transform the industry. Instead, it will be new business models reinforced by these technologies.
Read more: If machines can be inventors, could AI soon monopolize the technology?
Focus on the economy
The following figures describe a simple approach to focusing on economics, not engineering.
Figure 1 summarizes the basic economics of any investment decision. Invest if the extra profit is more than the “opportunity cost†– the benefit you can get by spending your money some other way or not spending the money.
Figure 1 can be difficult to use, so Figure 2 frames the investment decision in slightly more detailed terms using the economic concept of ‘marginal utility’ – the additional (marginal) benefit (utility) that comes from spending. additional.
To make this simple to apply, Figure 3 summarizes this decision making process into a simple “decision tree”.
The conversation / author provided, CC BY-ND
Solve Mike’s AI Investment Challenge
By applying this methodology to Mike’s situation, we can see why he couldn’t make commercial sense of the field of AI-based precision farming.
The seller answered the first question by stating that the gains from AI adoption would reduce Mike’s crop input costs by up to 80%. This would translate into a savings of about $ 80,000 per year for Mike (at best).
The seller also answered the second question, with a clear statement of the cost of the system.
But the business case failed on the third question. At best, the marginal benefit of adopting AI (savings of $ 80,000 per year) was just equal to the marginal cost ($ 80,000 per year) – not counting the initial installation.
Read more: Artificial intelligence is now part of our daily lives – and its growing power is a double-edged sword
Putting it this way clearly gives the impression that it is a failed investment, and one that it didn’t take long for Mike to speak out against it. But the point is, many decisions to invest in AI don’t make economic sense and the above process will make it easy to see why.
Using a value economic framework, rather than an engineering pretense of possibility, is the first step to making better decisions. This reduces the prospect of another AI winter and increases the chances of real gains contributing to a more prosperous and sustainable world.
[ad_2]
Comments are closed.