By Andrew Joseph, Editor

When it comes to AI, aka artificial intelligence, people either know all about it or don’t.

Every day, people assume that AI is a giant talking head of a computer that, after digesting all of the knowledge on the World Wide Web and digital social media, should be able to answer any question we pose.

That it can look and talk like a human is one thing. It might even innocently spew out a particularly harmful answer simply because it doesn’t understand social norms as well as some people do. Or perhaps it does understand them, which is why it spews out noxious dialogue.

Garbage in. Garbage out.

It’s true that on AI sites such as ChatGPT, we can ask the brain to tell us how it believes the yield and quality of the Saskatchewan wheat crop will be in 2024. Based on its current knowledge, it should offer a reasonable answer.

But that’s not the type of AI the agricultural sector needs to be concerned with.

Instead, we must concern ourselves with generative artificial intelligence, Gen AI or generative AI.

This is the technology where a computer program is fed as much information as possible about a local area—say the farm of one of your customers.

The generative AI software provides the farmer with the best option to proceed from all that data.

It’s not predicting anything; instead, it’s using the available data to provide feedback on how best to proceed in the near term.

A few companies use Gen AI digital crop advisors, which use agronomic data to turn it into actionable recommendations for farmers.

Analyzing big agronomic data provides AI-supported insights to optimize production practices. It helps farmers understand patterns affecting the performance of crop varieties and production on their specific farms. It tracks climate trends to help farmers become more resilient to the changing daily climate.

We discussed AI with Pearson plc, a company whose workforce solutions help ag businesses understand their workforce’s present and future needs, identify professional and skills development opportunities, and partner with employers and employees to future-proof their organizations in an evolving global economy. Headquartered in London, UK, Pearson also has offices in the US and Canada.

We talked with Jonathan Finkelstein, founder of Credly and the Senior Vice President of Workforce Skills at Pearson.

He explained what Gen AI is and how various sectors can use it.

“Until recently, organizations mainly used artificial intelligence for statistical analysis, processing vast amounts of data and offering outputs that would have taken humans far more time and energy to complete,” noted Finkelstein. “The advancement of deep-learning models, which allow AI to classify text and images and transcribe audio automatically, has given rise to Gen AI.

“Gen AI is a kind of artificial intelligence that—rather than analyzing or applying rules to data—produces something new: text, images, video, code, and other content,” he added.

Deep-learning models trained on massive amounts of data power Gen AI’s outputs. The AI model “learns” to generate a statistically probable output based on reconciling the prompt it receives against the data it’s been fed.

As Finkelstein explained to CAAR, Gen AI produces text relying on a large language model (LLM) trained on a high volume of text to develop patterns and probabilities of word choice and order concerning a given prompt.

“Much like smartphone autocomplete, models like ChatGPT generate words one at a time (very quickly), with each subsequent word based on the previous one.”

Models that produce code, images, and video also do so based on vast training data, influencing the outputs.

“In this way, Gen AI produces a new work that shows traces of its training data,” said Finkelstein. “When you ask a Gen AI model like ChatGPT to write you a song about Wednesdays in the style of Dolly Parton, for instance, it will use its data about the genre of a song, the style of Dolly Parton, and social commentary about Wednesdays to write what it deems to be the most statistically probable text that satisfies your prompt’s parameters.”

He said that Pearson believes Gen AI can positively impact how people understand and prepare for the changing world of work.

AI’s surge forward in our personal and professional lives has changed our perception of what’s possible. It’s also changing the reality of what’s necessary in the world of work. Thanks to Gen AI, tasks that have long seemed inextricably intertwined with particular jobs are now being cast in a new light.

“Our research (https://www.caar.org/wp-content/uploads/2024/03/U.S.pdf) focused on how generative AI will affect roles such as farm product buyers by automating or significantly reducing the amount of time spent on daily tasks,” related Finkelstein.

“Most generative AI models that have gained popularity are conversational AI chatbots, so rather than a “talking AI with a face,” these models use natural language processing to understand requests and questions and respond in writing.”

Finkelstein said that these models have the potential to automate repetitive tasks in the agriculture sector, which will make quick work of responsibilities such as scheduling and documentation.

Generative AI models trained on high-quality data can also offer data-driven insights and recommendations, allowing ag companies to work even more efficiently and use past experiences to inform projections like yield forecasts and product demand.

He continued: “Our research anticipates that within blue-collar jobs, farm product buyers will experience the greatest impact of generative AI, based on the tasks associated with the role and AI’s ability to reduce the time spent on those tasks. “With the right data sets and prompts, a generative AI model can use past buying trends, current inventory, and projected demand insights to make product buyers’ work more efficient,” added Finkelstein.

“By reducing the time spent on manual tasks like maintaining and reporting transaction and inventory data, generative AI will allow farm product buyers to use more of their uniquely human skills—communication and problem-solving, for instance.”

With generative AI utilized as an accelerator for repetitive tasks, farm product buyers will now have more time to spend on higher-value work requiring additional skill sets.

Pearson has developed data on AI. But where did it come from?

For the “Gen AI Proof Jobs” installment of their Skills Outlook series, Pearson used a combination of census and other workforce datasets to form a comprehensive view of the current workforce across the US, UK, Australia, India, and Brazil.

“Using our proprietary occupations ontology of 5,600 jobs and 26,000 tasks, we framed each job as a collection of tasks, allowing our machine-learning algorithms to calculate future technology impact on each job at a task level,” related Finkelstein.

He said that Pearson’s models weighed the future impact of 16 emerging technologies on each job’s tasks, considering the adoption rates for these technologies tailored by country and industry.

By projecting the percentage of time saved per task by 2032 and incorporating economic models and industry-specific growth patterns, Finkelstein said that Pearson could extrapolate which jobs Gen AI would impact the most. Or at least in those five countries over the next decade.

“Our “Gen AI Proof Jobs research found that Gen AI will have a more significant impact on white-collar roles over the next ten years,” Finkelstein pointed out. “Blue-collar roles—especially ones with more creative, manual, and collaborative tasks—are at less risk from the changes the rise of Gen AI will bring.”

He added that these insights are not meant to be considered alarm bells to scare individuals or organizations into making dramatic career shifts. Instead, they should drive opportunity and inform strategic workforce planning.

“Our research has uncovered a polarization between the tasks most impacted by technology—repetitive and technical tasks, such as scheduling appointments or answering and directing calls—and those less impacted, such as those requiring inherently human skills like creativity and collaboration,” noted Finkelstein.

“Looking to the future, the next questions individuals and organizations should ask include: How do we refine these human skills that future technology won’t replace and leverage technological advancements to augment our human work? And as we make progress, how do we provide continuous skills development, equipping people to transition into more valuable future, high-demand roles?”

Organizations that recognize AI’s potential for efficiency and scalability can begin offering employees more personalized learning solutions.

Finkelstein added that AI will empower people to make more valuable and productive career choices and be able to execute those choices more efficiently.

The company QuantamBlack, AI by McKinsey, revealed that less than a year after the debut of many Gen AI models, one-third of all its survey respondents reported that their organizations use Gen AI regularly for at least one business function.

“If the past year (2023) is any indication, we’ll see even greater advancements in Gen AI’s abilities and uses,” extolled Finkelstein.

A recent report issued by Bloomberg Intelligence predicted that the Gen AI market would surge by over 42 percent annually, reaching $1.3 trillion in value by 2032, up from just $40 billion in 2022.

Over the next ten years, we will also see an increased demand for specialized AI assistants, new infrastructure, and coding automation tools.

Major Gen AI companies will continue reducing AI-generated hallucinations—Finkelstein called these confident but factually incorrect responses—refine their training data and work to meet evolving regulatory requirements, such as the European Union’s recent announcement that it wants AI-generated content to come with a “warning” label.

According to Pearson’s data, farm product buyers would be the most impacted in the blue-collar job sector—some 27 percent affected per role hour at a task level. We asked Finkelstein just how Pearson arrived at the number.

According to Finkelstein, all of the models utilized by Pearson weighed the future impact of 16 emerging technologies on each job’s tasks, considering the adoption rates for these technologies tailored by country and industry.

The company extrapolated which jobs would be impacted the most (or least) by Gen AI by projecting the percentage of time saved per task by 2032 and incorporating economic models and industry-specific growth patterns.

“At Pearson, our purpose is simple: to add life to a lifetime of learning. We believe every learning opportunity is a chance for a personal breakthrough,” noted Finkelstein. “That’s why our (approximately) 20,000 Pearson employees are committed to creating vibrant and enriching learning experiences designed for real-life impact. We are the world’s leading learning company, serving customers in nearly 200 countries with digital content, assessments, qualifications, and data. For us, learning isn’t just what we do. It’s who we are.”

Summing up, Finkelstein reiterated how Gen AI can help the agricultural industry.

“Pearson’s workforce solutions can help agricultural businesses understand their workforce’s present and future needs, identify professional and skills development opportunities, and partner with employers and employees to future-proof their organizations in an evolving global economy.”