In just over a year, artificial intelligence (AI) programs like ChatGPT, Midjourney, Bard, and more have gone from a novelty to a viable tool for an array of industries – and the seafood industry could capitalize on them to increase sales and target more customers.
The use of AI is not new to the seafood industry but typically is relegated to uses in the aquaculture industry. An analysis by ThisFish CEO and Co-Founder Eric Enno Tamm – shared during Seafood Expo Asia in September 2023 – found the vast majority of AI use in the seafood industry takes place in aquaculture. Tamm said most of the investment is in aquaculture as well, and that the uses of AI are typically relegated to machine vision technology – programs that use machine learning and AI to analyze video or images.
Those programs analyze images to generate data for companies – in most instances, the machine vision analyzes how fish swim and allows companies running aquaculture operations to optimize feeding times or identify problems before they arise.
Native AI Co-Founder and COO Sarah Sanders – speaking during the Global Seafood Market Conference (GSMC) in Orlando, Florida – said her company has been in the AI business for over six years and focuses on the consumer side of an array of industries. That's
“Google search and the ever popular ChatGPT and other conversational AI tools are all powered by NLP, and what it does is it takes large masses of text, it reads it in fractions of a second – something that can take days, weeks, months for us to do internally the old ‘manual’ way,” Sanders said.
Those language models can then analyze that text and pull things “like sentiment, you know happy, delighted, sad, angry, afraid, whatever that might be,” Sanders said.
It can also analyze text and identify the frequency of certain phrases quickly, pulling out specific phrases or suggestions to determine how many times something is mentioned.
“And we train algorithms to do this quite quickly,” Sanders said.
By doing so Native AI and its team have created a number of different algorithms that can analyze wide swaths of data. For GSMC, the data Native AI used was publicly available reviews of an array of seafood items posted on some of the largest retailers in the U.S.: Costco Wholesale, Albertsons, Sam’s Club, Target, Walmart, and Whole Foods. Native AI then used an NLP to analyze 20,000 reviews of popular products to determine the top issues and common perceptions people had. The data revealed seafood shoppers are more satisfied with the price of the products than shoppers of other products like prepared chicken.
“We’re not looking at numbers, please keep that in mind. We’re talking about price satisfaction and written words in the context of how people are leaving reviews,” Sanders said. “We’re not doing a price comparison, because that’s not something that’s easily done.”
Another takeaway from Native AI’s research was that seafood consumers leaving reviews said vacuum-sealed products, like those using skinpacks, were the best in terms of freshness and quality.
“The average consumer is going to have a certain expectation like, ‘I am buying fresh seafood.’ So when it is not exceptionally fresh or at least as fresh as they thought, especially when it’s expensive, it’s really disappointing,” Sanders said.
Canned seafood products were also ranked high in terms of perceived freshness, which can be attributed to the gap between expectation and reality, Sanders said.
Another piece of Native AI’s research found 84 percent of seafood consumers were very satisfied with the price of prepared seafood items, compared to just 48 percent of prepared chicken shoppers.
Sanders said the key takeaway from Native AI’s work is that NLPs and other forms of AI are available ways to determine what consumers are already saying publicly, without requiring any detailed surveys or market research.
AI can also be used if a company wants to do that market research.
The NLPs also allow companies to extract data that is gradually becoming less available as more rules come out about tracking consumer habits online – like the General Data Protection Regulation (GDPR) passed by the European Union.
“There are data points with natural language processing that you can extract really quickly to at least have a clue,” Sanders said.
Using that, Native AI has created a tool it calls “digital twins” that allow it to create the persona of a virtual consumer, which can then be used for market research without requiring expensive surveys.
“We can ask them questions, these digital twins. Like [asking] Kroger customers, 'What’s the biggest factor when selecting a meal?'” Sanders said. “Competitor customers, you can ask them questions as well.”
The digital twins are again created using NLPs, using publicly available data, to give companies a better idea of who is shopping for what, and why they're searching out those products. That data can in turn be used by companies to either develop new products are better position existing ones.
“These are just some examples of where AI is going and the innovations ide of the house with NLPs,” Sanders said.
Photo by Chris Chase/SeafoodSource
