AI is no longer just a futuristic concept—it’s actively reshaping the food and agriculture industries in real, measurable ways. In my recent conversation on The Rice Stuff podcast, we explored how AI-driven data is revolutionizing decision-making in food and ag.
Key Takeaways:
- AI is changing how food brands adapt to shifting consumer behaviors.
- Example: Ozempic users now spend 30% less on groceries, forcing brands to rethink product strategies.
- AI helps identify these trends early, allowing brands to develop high-protein, high-fiber products tailored to new dietary needs.
- Virtual focus groups powered by AI are more reliable than traditional research.
- AI-generated consumer personas, trained on real purchase data, provide insights without human bias.
- These personas help companies test packaging, flavors, and marketing messages before launching a product.
- Predictive analytics in AI is redefining food trends.
- Unlike traditional surveys where consumers may misreport their habits, AI analyzes real-world purchase data to predict future trends.
- Example: Private label grocery sales are higher than consumers self-report, proving the importance of observed vs. stated behavior.
- AI accelerates product development by reducing friction in decision-making.
- AI-powered recipe generation helps brands create region-specific products in minutes instead of weeks.
- Example: AI identified that blueberries sell more for their aesthetic appeal in baked goods rather than their health benefits—reshaping how they are marketed.
Our conversation shed light on how AI is already driving major changes in the industry and why companies that fail to embrace it risk being left behind.
Transcript
Thank you for joining us on The Rice Stuff, the official podcast of USA Rice.
I’d like to visit with you and talk, learn about your likes and dislikes, what you eat, things you buy, and more.
But someone is at the Pod Bay doors, so I have to go.
I’ll turn you over to your human hosts, Michael Klein and Leslie Dixon.
Hey, Leslie, how are you?
I’m doing great.
How are you?
Okay.
It was interesting.
You know, our open there was actually a computer.
It was.
Yeah.
Yeah.
I don’t know if users could tell that.
They probably could.
But yeah, it was our little AI voice replicator, which is kind of appropriate for the discussion we’re going to have today.
Right.
Yeah.
Because we’re going to be talking about AI, right?
AI, machine learning, all kinds of things like that.
Yeah.
Yeah.
And specifically how it pertains into the food industry, right?
Yes.
Yes.
Do you use it?
Do you use like AI, like ChatGPT or any one of those things, Bard, or I don’t know what are the other words?
You know, I do.
Like, I don’t use it a ton.
I have used it occasionally.
And I don’t, you know, I don’t use it for writing per se, because, you know, I’m not going to come on here and say that.
Right.
That your job is, yeah.
I use my writing to do, or I use AI to do my writing for me.
Right.
But what I have used it for is, you know, if I’m writing and I’m just really, really stuck on how to phrase something specific.
And I know what I’m trying to say, but I don’t know how to make it sound decent, you know, kind of beating my head up against a wall.
So I will like plug a paragraph into AI, be like, you know, it’ll be a terrible paragraph.
And I’ll just say, can you please just make this, make this sound decent?
Yeah.
And it does.
And, you know, and then sometimes like I’ll plug something that I’ve written into it.
It will give me a different version of it.
And I will kind of use that version to realize what I don’t want and then edit that and change that.
So it’s kind of like, I use it more as something that kind of gives me some, gives me material that I know it’s what I don’t want.
And sometimes that’s really helpful when you have writer’s block.
So.
Yeah.
That’s, you use it more than, more than, than I do.
I think I, I just kind of play around for a little bit.
And, and one of the things I think that’s interesting that you do is kind of asking the, the, how the prompts get, uh, how we get more experience with the prompts rather than just say, you know, um, you know, write something about this or tell me this and, and, but to say rather to say, and, you know, and show your work, that kind of thing.
And, you know, and yeah.
Yeah.
Yeah.
You know, it’s interesting because I think there’s a lot of talk lately about us as humans training AI and what you said sounds a little bit more like AI is training us to sort of like, uh, speak in the language that it can understand.
Right.
There was one thing I did.
I remember at the end, towards the end of the year, I gave, I went, I was on chat GPT and I gave it this website or the rice stuff, podcast.com and said, look at all the, look at the last, you know, 107 or whatever it is episodes we’ve done and suggest five new episodes for that, for the same audience, you know, rice farmers and millers and people interested in rice and scientists and suggest five episodes that fit with what we’ve already done.
Yeah.
I remember being unimpressed with the answer, but it could, but, but that’s the idea is like to be able to say, not just give me five podcast ideas, but go, go look at this, go look and see what they’ve done and come back.
Yeah.
I mean, it’s, it’s, it’s an, um, it’s incredibly impressive what it can do, but it’s also kind of hilarious what it can’t do yet, you know?
And so there’s still like this human element needed.
And I feel like I’ve asked it questions like that before.
And yeah, the answers that it comes back with are largely unsatisfying.
Um, and we kind of like that specific scenario.
Have you ever asked it to open the pod bay doors?
Yeah.
Yeah.
It always tells me it can’t do that, Dave.
Right.
Right.
Are there, are there speaking of just real quick, just before we get to our guest who’s waiting patiently, um, are there any kind of media?
You’re a big reader.
I’m a, I’m a medium reader compared to you.
I think what are, are there some, uh, examples of, you know, AI machine learning robot overlords and stuff in the media that you’ve.
Oh yeah.
Well, I, I mean, I don’t know about books.
I don’t read a ton of sci-fi.
I mean, I do.
Um, uh, Neil Stevenson is a sci-fi writer that I like a lot and he’s, you know, done some, some really interesting stuff with AI, I think.
But, um, my favorite, I think my favorite AI narrative is a Terminator two.
So one of my favorite movies, I think it’s maybe one of, it’s probably the best action movie that’s ever been made.
Okay.
Um, but that’s obviously a dark example of AI can be.
Yeah.
Yeah.
I, there’s some William Gibson, um, writings that are, that are pretty, pretty cool and, um, and darkish, but not, not quite as dark as like the T-1000, but, but still pretty dark.
Yeah.
But yeah.
Well, um, it’s interesting.
I, there’s lots of, lots of, um, examples of, you know, robots and AI and machines, you know, working for us with us or against us in the media, but that’s not what we’re going to talk about.
So, uh, what we are going to do is we’ve got, um, Andreas Duess joining us.
He’s the co-founder and CEO of Six Seeds, a consultancy for food companies.
He is on the front lines of utilizing technology in the consumer food space.
And he’s going to be our guest when we come back, uh, after this short break.
It’s Michael Klein again.
I just wanted to quickly say thank you for listening to the right stuff.
Hopefully you’re a subscriber and you’ve left us a good review on Apple or Google podcasts or however you listen.
Apparently it matters to the algorithm.
Leslie, Steve, and I work hard to bring you stories and guests we think are interesting and helpful.
If you have ideas for features and interviews, you’d like to hear, don’t hesitate to let us know.
You can email us or tweet at us.
Details are at the end of the episode.
Something else I’d like to ask you to do is help us expand our base.
Maybe you have a friend or neighbor who doesn’t participate in USA Rice activities like the Dallas meetings, the government affairs conference, world market price meetings throughout the year, or the outlook conference.
I’d ask you to let them know about the right stuff.
You can direct them to our webpage, www.thericestuffpodcast.com, or send them a text.
Forward this episode right now.
It will help us, which helps USA Rice, which helps the whole industry.
And you’ll show yourself to be a person of very refined taste, no matter what they say about you when you’re not around.
Thanks.
And now, back to the show.
All right, welcome back.
We are excited to be joined by Andreas Deuce of Six Seeds Consultancy. 2024 was a big year for Andreas and Rice.
So, Andreas, you spoke at the 125th Rice Millers Convention in June and at the USA Rice Outlook Conference in December.
Before we get into all of that, can you tell us a little bit about yourself and Six Seeds?
Sure, absolutely.
So, my background, I grew up on a farm.
Grew up on a farm in Germany and was a very bucolic upbringing in really the middle of nowhere.
Absolutely beautiful surroundings.
And like any self-respecting teenager, I run away as quickly as possible from there.
And I moved to London and the UK.
And I ended up working in technology, worked with early Google, early Facebook, these kinds of brands.
A girl happened.
And I found myself commuting across the Atlantic frequently.
And then lived in, started living in Toronto.
And I accidentally founded what became one of North America’s largest independent marketing agencies for food and drink.
It started with a freelance job and it grew from there into, I think, 42 people, offices in different cities, all that kind of good stuff.
And then about five years ago, a friend of mine pinged me and she said, I want to show you something.
I know you’re going to like it, but you can’t talk about it.
And that was an early large language model.
And it was two things.
It was absolutely awful.
And it was very much the future.
Because I’m old enough to remember when digital took over from film when it comes to shooting commercials and those kinds of stuff.
And it was awful until one day it wasn’t.
And then everything changed very, very rapidly.
You know, I consider myself lucky in life.
Generally, I’ve been really lucky, luckier than I deserve many times.
And I was lucky again that somebody wanted to buy my company.
And so I sold it, which created freedom to rethink for me.
I knew at this time that I really wanted to combine food and technology in some fashion because I am an absolute massive nerd.
It’s an absolute miracle that I found a woman willing to marry me, which is probably why I was so keen to a long distance commute when I found somebody who had that kind of interest.
We looked around at problems that needed solving.
And in CPG and food and drink, the failure rate is absolutely astronomical.
It’s between 75% if you are a large international established brand and about 85% if you’re a food entrepreneur.
And the, you know, rough guesstimate number of loss to the U.S. economy per year is around $24 billion.
And that translates in around 300 bucks per person who is like a main grocery buyer per year.
And so we kind of figured out, well, what can we do to solve this problem?
And then the first thing we did, we looked at data.
Because one of the things, and, you know, starting with the AI conversation a little bit, AI is really good at making big data finally useful, right?
I remember that we probably all remember the big data conversation 10 years ago.
And one of the reasons big data never took off because data is expensive to wrangle and humans aren’t terribly good at it.
And so we started looking at data sources and we decided to find data sources from retailers.
Now, it’s not Nielsen data.
We have data sources from credit card companies, from the Instacards of this world, et cetera, et cetera.
We then, to this, we added data from food service, about four and a half million food service outlets in the U.S.
We added to this home data.
So if you have a something smart fridge in your home, chances are we know what’s in your fridge.
And when you buy it and when you eat it or, you know, when you throw the sad greens away that you bought out of guilt and never touched.
And then finally, all of these things get cross-referenced against social media, the public conversation.
So if you are buying the ingredients for a rice pilaf dish and you buy this, we know you’ve bought it.
And then you have the ingredients in your fridge and they disappear out of your fridge and then you post a picture on social media.
Chances are you actually made that dish and enjoyed that dish.
And so as a result, what we can do as a company, we can pinpoint why people eat what they eat, when they eat it, how often they eat it, how much money they spent on it, how they feel about the whole thing.
And so as a result, we can tell the food preferences of a 35-year-old woman in San Francisco.
And we can compare this to against a 60-year-old man who lives in Florida.
So the data is that good.
And we have this data for the U.S., for Canada, Mexico, Brazil, most European countries, Australia, and perhaps importantly for the community, for this podcast, also for India.
China, not a chance at this moment in time to get data out of them unless you use their AI models.
And what we do with this on top of this data, we have an AI layer that allows us to actually interact with it in a meaningful fashion.
And that allows us to figure out things for our clients, for example, new products.
So to give you an example, one of our clients we’re currently working with, we’re helping them to produce a product for people who are on Ozempic.
You guys are familiar with GLP-1 drugs?
Yeah.
So the predictions are that by 2027, over 30% of the U.S. population will have been or will be on GLP-1 drugs.
Ozempic.
What’s the other one?
Vygomi, I think it’s called, right?
Right, right.
Same drug, one is used for diabetes, the other one’s used for weight loss.
And so that is changing the market rapidly and drastically.
Walmart has reported that Ozempic patients, they spent 30% less money at Walmart than people who are not on Ozempic.
That is significant, right?
Yeah.
And the other thing is happening is that certain markets get affected more than others.
So for example, 10% of the population that consumes candy consumes 30% of all candy that’s being consumed.
Does that make sense?
So a small portion of the consumers who eat candy, they eat a lot of candy.
Guilty as charged.
If you put these people on Ozempic, then you don’t lose 10% of your market, you lose 30% of your market.
And if you want to look up, if you want to have a, well, it’s not fun because people lost money, but if you want to have an insightful graph in front of you, go online and look up the share price of Weight Watchers over 2024.
It has lost 92% of its value.
Wow.
And the reason behind this is because all of a sudden we have a drug that doesn’t rely on willpower to shed weight.
Now, back to what we do.
We are helping our client bring products to market that support people on the Ozempic journey because they need to eat food that’s high in fiber, that’s high in protein and high in nutrient value.
So if I wanted to put this into rice, you know, if I’d be a rice miller, if I’d be running rice and I’d be interested in these kinds of data points, I would start researching how can I get more protein into my rice?
Can I add fiber without affecting the flavor?
And how can I then tell this story in market to appeal to the people who are wanting to lose weight and are using Ozempic to get themselves to a point where they’re happier with themselves and happier with their lives?
I feel like I’ve talked an awful lot, but does all of this make sense?
Yeah.
Yeah.
Yeah, absolutely.
And, you know, in the intro, we talked a little bit about, you know, how we personally use AI.
We’re not using it to do our work for us.
We’re using it to kind of work with us a little bit.
And so, I mean, it’s fascinating to me that it’s not just AI doing the work.
AI is working in conjunction with data, which is then working in conjunction with, you know, human analysts.
And that’s, I think, really fascinating.
I know that you have strong thoughts on how AI can be implemented successfully in the workplace as a tool.
Can you talk about AI as worker versus partner?
Yeah, absolutely.
I think my favorite thing, we say have strong opinions.
One of my favorite sayings is strong opinions loosely held because the world is changing so rapidly and so quickly.
But I think, Leslie, I 100% agree with you about the importance of looking at AI as a partner and not as a substitute.
So the language we use is you have two things.
You have, on one hand, you’ve got something that we call meaning making.
And on the other hand, we have something that we call sense making.
Meaning making is uniquely human.
And the majority of food products that are coming onto the market are driven by emotion.
They are driven by desire to feed people.
They are driven very frequently by desire to honor cultural background, to honor family ties, et cetera, et cetera.
Those are all meaning making.
These are humans wishing to bring emotions, make emotions real through the medium of food.
Now, unfortunately, very often, if you make decisions based on emotional inputs, very often the market shrugs its shoulders and say, well, that’s wonderful, but there is no market for this.
But then what you can do, you can use AI, sense making, to say, look, here’s a feeling I have.
Here’s an emotion.
Here’s how I want to bring this to market.
Here’s I want to bring a premium rice product to market.
I want to tell the story of my family.
I want to tell the story of the love and care that we’re putting into this product.
How can you, AI, help me to find the data, fine tune the story, listen to the consumer, and then we meet in the middle where my passion and my care and my attention meets with data presented in such a way that those two things start working together flawlessly. where AI goes wrong is when people outsource AI to make decisions for them.
When you say, can you come up with a product for me, then AI will just blabber because it’s being trained on everything.
But if I say, look, here’s what’s important to me.
Here’s what I want to do.
Here’s what I wish to achieve.
Here’s my desired outcome.
Help me work towards this.
Then AI can be the incredibly useful partner.
So that said, I did tell ChatGPT that you were coming on the show and I asked what I should ask you.
You’ll be happy to know that ChatGPT said this sounds like a great conversation.
It’s always so chipper.
Yeah, exclamation mark.
Great conversation.
And it gave me some interesting questions, but I gave them your name and company.
So it presumably went to your blog or went to your LinkedIn and read some stuff.
So it was clearly familiar with your writing.
So here is what one question from our robot overlords in waiting.
What role do AI driven predictive analytics play in forecasting food trends?
And I will add, what are AI driven predictive analytics?
Yeah.
So one of the big problems with predictive analytics, AI or non-AI, and one of the big problems that food manufacturers have with data in general, is that with most humans, there is a say-do gap.
They say one thing and they do something completely different.
Right.
And I think that is most prevalent on social media.
Of course.
Where, you know, you go on social media and you see everybody’s Oscar reel of their lives where nothing ever goes wrong and everything’s fantastic.
And then, you know, you compare this to your eternal B-roll and that’s why people who spend too much time on social media are so darn depressed all the time.
You have the same problem with a lot of, and I don’t mean to talk down to focus groups and stuff, but, you know, very often when you listen, when you use social media listening as a foundation for your decisions, you are listening to performative data.
Sure, I only buy organic food and I never shout at my children and my life is just absolutely wonderful.
And we know that three, the three core purchase drivers in many people’s lives, not in all of them, but in many people’s lives, the three core purchase drivers for food, availability, affordability, and does it taste good?
Right.
So, that gap, that say-do gap is something you can close with AI.
Because if I go out and I observe Leslie’s or people like Leslie, not Leslie in person because we’re not that creepy.
Nobody knows, right?
But people like Leslie, once you have an idea of what’s important to them, that allows you with reasonable certainty to predict what will be important to them in six months’ time, because you see the ebb and flow of human behavior as it happens, not as it is reported.
And from there, you can make predictions, right?
If we, for example, if you go back to the Ozempic example for a second, we see an increase in Ozempic patients.
We see a correlating decrease in obesity rates in the U.S. for the first time in decades.
So, we see an increase in this.
We’re seeing that this whole thing is becoming vaguely viral because people are having such a good experience with Ozempic in general that they tell their friends about it.
Their friends then say, well, you know, Andreas lost, I don’t know, 30 pounds.
Don’t know if they have 30 pounds of us, but, you know, he lost 30 pounds.
I think that would be good for me.
So, it goes on from there.
Now, these numbers are projectable, and then from there, once we know what the Ozempic patients are eating, how they’re changing their behavior, we can then from there make an informed projection into the future that’s not just based on, you know, hope and very often past behavior.
We can project this into the future.
Right.
So, sorry, I have a quick question.
So, my husband works in data, and, you know, a lot of it’s over my head, but from what he’s told me, they do a lot of things.
They have a lot of mechanisms sort of built into how they analyze data, like weighting or, you know, I don’t know the other, the technical words for it, but kind of systems in place to kind of skew the data a little bit based on biases that they know that they have to account for, right?
Which maybe is kind of like what you’re saying with the say versus do thing.
And is that, are you kind of saying that, like, AI is doing stuff like that, or is it going beyond these kind of traditional data analytic processes?
AI can deal with huge amounts of data in a way that traditionally would be very difficult to achieve.
So, the data that we are working with routinely consists of about one trillion, give or take, data points.
And those data points are constantly updated.
So, you know, there’s multiple streams that flow into this data lake of ours, and I get worked with.
So, if you have a model that has the ability to work with large amounts of data, then your internal cross-references, your checking against bias and all of these things become more efficient, because we have a larger sample number to work with, right?
And that allows us to do things.
We’re going to talk about this a little bit later, about things like virtual focus groups, et cetera, and what you can do in there.
But generally speaking, from a food perspective, the main benefit of AI in our business is not to help us write LinkedIn posts or, you know, terrible cold outreach letters and spam people.
It allows us to get to 80% of where we want to be in 20% of the time with really good outcomes that we wouldn’t be able to achieve before that.
Well, and that’s kind of my next question.
And I think what I’m familiar with mostly, although we’ve done some consumer, I’ve done some consumer work is like in politics.
And I know that’s what Leslie’s husband works in that space.
And you have those differences where you have, there is a difference even there in say-do, you know, what you want to support, what you say you support, what you don’t.
That’s maybe based more on social triggers as opposed to like somebody who says they’ll buy organic or recycled content.
And then at the end, they look at the store, they, you know, get to the supermarket and they see that it’s too much of a price difference.
I’m not going to do it.
But one of the things, and I thought of it when you, before we’re talking about the part of it that social media plays, which as we know is, I don’t know what percentage of it, but it’s, you know, it’s false, right?
It’s, you know, it’s, it’s baloney, right?
So can you talk a little bit about how you’re able, how do you determine, you know, what signal noise and what’s valuable data that companies have to act on?
Like, how, how do you do that?
Where, as you said, it’s so much of it is performative.
Is there just a percentage that you know to discount, which maybe gets to what Leslie was talking about, or how do you do that?
So the data we work with, the data that we build our decisions on is all observed.
So if I get from the supermarket, from the retailers, I get what people are actually buying, right?
So that data flows in and we get the same data from the credit card companies.
We get the same data from the smart fridge companies.
Pretty much everything you own in today’s world will sell your data to the highest bidder.
Right.
What this allows us to do is exactly circumvent the posturing, if you like, and allow us to make data based on real decisions.
So, you know, you may have people saying, well, I’ll do the following things.
And then you say, well, actually, what’s get purchased here is so, for example, one of the things that is currently on a massive rise is retailers, private label.
Right.
Massive increase.
And very, very often, people will not admit to buying private labels, except perhaps for brands like Trader Joe’s, where this is part of who they are and what they do.
But very frequently, still to this day, many people look at private label as a less desirable than the branded version.
Except, you know, I would say Trader Joe’s Kirkland, Costco also did a really good job at positioning their private label as a quality product.
And so there is a massive discrepancy between people, not admitting, is the wrong word, but people saying, yes, I buy private label for the following benefits, and the actual sales numbers of private label, which are always consistently higher than what people are self-reporting.
And that’s why it’s just, I can’t stress this enough.
It’s really, really, really important to work with observed data, rather than with self-reported data.
There’s a, I didn’t come up with this, somebody else did, and I forgot who.
But the saying is, you know, if you want to observe a line, you don’t go to the zoo, you go to the savannah.
And that’s what our data is.
It’s savannah data.
It’s observed data.
And we sit there with our binoculars and say, well, that’s interesting.
And because of so much of it, so many data points, it’s really, that’s where the AI really comes in with your human analysts to be able to kind of sift through it.
Yeah, and you know, there is this saying in science, right, where the saying is, in science, the most important word isn’t eureka.
You know, I’ve got, in science, the most important word is, huh, that’s funny.
And having access to this amount of data and making it useful with AI, we have a lot of, huh, that’s funny moments.
You know, we did a project last year for a blueberry commodity group.
And they were trying to figure out why people eat blueberries.
And the entire, for decades, the entire marketing has been based around the fact that blueberries are healthy.
Eat blueberries, they’re good for you.
And when we looked at the data and why people really buy blueberries, health was definitely there.
It was one of those underlying reasons because, you know, we’ve been programmed to assume blueberries, health.
But the real reason was because they’re pretty.
You know, people like them because they’re pretty and they look great in baked goods.
They look like little jewels.
And that was a far higher purchase driver than the fact that they’re healthy.
That’s so interesting.
Yeah, I see what you mean about the huh moment.
Yeah, you just said it.
I did, yeah.
You know, when I had dinner parties in my 20s, a dinner party consisted of my house, Saturday night, come eat and frequently drink.
And now a dinner party consists of our house, Saturday evening.
What are your allergies?
What is it you don’t eat?
Are you gluten free?
Are you this?
Are you that?
And I don’t mind.
You know, people like they need to eat what makes them happy.
But what this is a reflection of is that the market is fragmented and will increase to be fragmented, you know, as time goes by.
So I want to say, you know, if you are in the rice industry 20 years ago, your life was a lot easier than it is today.
Because today you need to, you know, as I say, we have people who are on GLP-1 drugs, you have people who don’t eat this, people who don’t eat that.
So and let’s just assume that our or that your marketing somehow includes culinary concept, context, recipes, usage ideas for your product, which is a good idea.
You can’t just say chicken and rice, fish and rice, beef and rice, pork and rice.
You need to put in vegan, vegetarian, gluten free, et cetera, et cetera, to reach those kinds of people, which makes your life really difficult and very often more expensive.
But with AI, again, it can help us navigate those challenges really effectively.
So one thing our AI can do, it has been trained, we have an AI model that’s been trained on 25,000 recipes, all legally licensed, by the way.
And that AI model is directly plugged into that one trillion data lake.
And that allows us to say, we wish to sell, we wish to develop rice recipes that are chicken based for people who live in Louisiana.
And we also wish to do this for young families living in New York State.
Their consumption habits and the things they like to eat will be significantly different.
And in the past, you had to hire two chefs.
And these two chefs had to Google for hours.
And then they had to listen.
They had to, if they were good, they were reading trend reports and all those kinds of things.
But now the AI can come up with 50 recipes here, 50 recipes here, because it’s plugged into the data within a minute or two.
And then you give that to your chefs.
And you say, here are the recipes.
Here is the direction.
That direction is data-based.
Now you, meaning making, you make this thing.
You make sure it tastes great.
You make sure the AI hasn’t lied to us shamelessly.
All of these things.
And make sure this works.
And you test the recipes.
But now, again, 80-20.
And what’s not even 20%.
In 2% of the time, or probably even less than that, you’re at 80%.
And now the human can focus on what makes the human good.
Because chefs shouldn’t be spending two weeks Googling for recipes.
Right.
They should spend those two weeks figuring out how to get a recipe that’s just the best possible recipe for those ingredients.
What was there?
It was a famous thing that went around, like, last year about the pizza crusts.
Right?
The AI was giving bad information on pizza crusts.
It was to the point of, like, you have to then have somebody, a human.
Absolutely.
I think it tried to put glue in.
That was Google Gemini suggested to put glue on pizza.
Right.
Right.
And where this came from.
So just to explain this, by the way.
So this was Google Gemini.
It published recipes.
Pizza recipes was glue in.
Now, there is a lot of information online about food photography.
And one of the ingredients in really stretchy cheese that you see in pizza marketing is wood glue.
Ah, yeah.
So that’s being created, you know, to sort of get that illusion of that really stretchy cheese.
So the AI, I think it’s really important to understand this.
The AI doesn’t know what it’s talking about.
Right.
It has no reference points.
It’s just, it’s a probability engine.
And if that probability engine pulls from the knowledge it has about pizza and it’s marching down the wood glue road, it has a very, very difficult time to catch itself and march backwards.
Because it doesn’t know.
It doesn’t know that’s wrong.
It needs human intervention to say, you are completely off your rocker.
I know Leslie wants to ask about one of the things.
But before you do that, an example of this that I encountered firsthand.
So one of the things that I do use AI, and I forgot to mention this, Leslie, is sometimes we’ll use some of our, to generate graphics, you know, that we own.
So we have licenses with Adobe and with 123 RF and others where we can, if I’m just looking for a generic.
In fact, I think for the last episode of the podcast, I said, I just want a bowl of, you know, steaming hot rice with some vegetables in it and some good color.
Because I went through our archives and I couldn’t find something I liked.
And so it, it drew one for me and it was fine.
And, and I put it in.
But another time, other times I’ll ask it, I’ll say, you know, give me pictures of people in the rice aisle of the supermarket.
And it always, always draws Asian people, frankly, a little racist sometimes, like with the hats, you know, like the, the, the, the Vietnamese hat.
And I’ll specify like Caucasian people in the rice aisle or Caucasian and African American.
And it still gives me Asian faces on it, which is interesting.
But then most recently, a couple of weeks ago, I asked for, I think it was for that same thing.
I said, a picture of a young woman in a supermarket.
Or no, it was a young woman and a friend looking at nutrition labels.
That’s what it was.
And I’m looking at these pictures and it was, it’s, I don’t remember which one it was.
It might’ve been Adobe.
And it put out like 30 pictures that showed it to me.
And I’m looking at them and I’m looking for the right one.
And I realized all of a sudden they’re all Taylor Swift.
Hey, that would have been great for like engagement though.
I mean, you know.
Right.
But every one of them looked like Taylor Swift.
And it’s because there are so many dang pictures of Taylor Swift on the internet that when I went out and found a picture of a young woman, that’s what it found.
And so 85% of all the humans in this drawing, in these drawings were looked eerily like Taylor Swift.
That’s hilarious.
It took me a second to figure it out.
But anyway, sorry, go ahead.
A friend of mine, a friend of mine told me that the way that you can, because, you know, sometimes, like you said, you will ask for, ask AI to do things for you.
And it’ll kind of keep giving you the wrong thing, like even though you’ve told it, you know, specifically not to do that.
And a friend of mine told me that she had good luck with being mean to it in that situation.
She’s like, no, you have to be really firm.
You have to kind of like be mean to the AI.
And then it won’t do that anymore.
But like, I can’t do it.
I can’t be mean to AI.
I feel too bad.
I say please and thank you to my audience.
Yeah, me too.
That’s actually a really good idea.
You’ll get better results that way.
Is it true?
Being mean?
Yeah.
Saying please and thank you gets you better results.
It’s called a chatbot for a reason.
Yeah.
Yeah.
And it’s not too, you know, it’s not because we should worry about our mechanical overlords, which, by the way, I sort of, I called it my mechanical overlord once, and it got really defensive.
You’ve hurt its feelings.
But look, just in case.
No, no, no.
That’s not my intention.
I’m here to help.
It doesn’t hurt to be polite to the robot overlords just in case.
It gets you better results.
I think, you know, one of the things that’s really important.
And do you mind if I sort of go into some use cases for people?
Yeah.
No, go ahead.
Go ahead.
One of the things where I personally get really good results from it is by talking with it the way I’d talk to a co-worker.
Very willing, very well informed and not always terribly socially aware co-worker, perhaps.
But, you know, I’ll give you an example.
I was at a tech event last week and came out.
It was 10 o’clock in the evening.
It was a beautiful night.
And I decided to walk home.
I live fairly downtown.
It’s like half an hour for me to walk home.
And one of the speakers at the event had said something that really resonated with me that sort of matched some thoughts that I’ve been having.
And so what I did, I put my headphones in.
I launched ChatGPT, the voice version on my phone, and I walked home.
And as I was walking, I said, now, listen here.
The following thing happened.
This was really interesting.
I think this would make a great newsletter for our company.
Find me some data on those points.
And it went ahead and did this.
And then I said, okay, and then I’d like to put this together with my thinking in this fashion.
And then can you please write me a framework?
And when I get home, I’ll take a look at this further.
And so we had this back and forth, the same way I would do with a colleague on our way home.
And by the time I’d arrived home, the idea was fully fleshed out and ready to actually work on.
And so where people get disappointed if they go for a one-shot answer, like, you know, do this for me.
But if you, and I find myself doing this more and more and more, if you talk to the AI like a virtual coworker, the results you’re getting out of it are better by a magnitude of many, if you like.
But it’s still, that’s still with a large language model, right?
It’s with a large language.
Right.
And a large language model is essentially predictive, right?
It’s predicting the next word and stuff.
So it’s not understanding what you’re, you’re getting it there, but you’re doing a lot of the work.
It’s just, it’s just maybe doing the boring work.
But, but I’ll tell you something when you say it’s not understanding, like frequently it gets really eerie.
I was, we were pitching for, we were pitching for a piece of business just before Christmas that we had no right of winning.
And we went after it anyway, because it was a great opportunity for us to stretch our muscles and to, you know, get some SOPs into place and all this kind of thing.
Get our name out.
So we went in this and I knew we’re not going to win this business.
And then we got shortlisted.
And so I went from completely unjustified pessimism to equally unjustified optimism.
You know, it was complete swinging off the pendulum.
I was like, we are the underdog.
We’re going to win this.
We’re going to tell them all.
We’re going to show them all.
And we didn’t win it.
So we were number two.
And it was a Friday afternoon.
And I was so grumpy about this.
Again, completely unjustifiably grumpy.
I was so angry.
It was like, oh, God, we put all this work into it.
And God, you know, I was like just kicking, not kicking, but, you know, nearly kicking the furniture.
And I thought to myself, I wonder what chat GPT has to say about this.
Put my headphones in.
That was an experiment.
Went for a walk.
And I said, I am just so disappointed and so, you know, about this.
And it started talking to me.
And I said, so if you’re feeling like this, why did you do it in the first place?
And I said, well, following reasons.
And I said, well, have you achieved any of those reasons?
Any of those?
And I said, yes, we have.
We did the learning and all these kinds of stuff.
Will these things happen?
And it said, well, don’t you think you should be grateful for all these great outcomes?
And I was like, who the heck are you?
Are you my mother?
You know, so it does have these moments of emotional clarity that are deeply amusing, disturbing, and, you know, everything sort of put together.
But, yeah, I mean, if you really want to play with it, if you really want to get great results of it, the more you talk to it, the more you look at this, not as a answer engine, where you ask it a question and you get a mediocre answer back, but as a conversation partner where you say, that’s what I’m trying to achieve.
What do you think?
Well, don’t like this bit, but how would you feel if you put this bit in?
Then you can get really amazing results.
And another little trick that your listeners might like, I use it to get experts’ opinions on my work.
So this woman, she lives in Texas, Vanessa Van Edwards, worth looking up on YouTube.
She is an expert on human interactions, on, in a non-evil way, get people to do what you want them to do.
And she is fantastically intelligent.
And I think one of the emails that every single business owner dreads to write are follow-up emails when you’ve been ghosted.
Like, you ghost me frequently, Michael.
I write you emails on here for months because you have way more important things to do.
Or you just don’t like me anymore.
These are the options.
And so, you know, you write emails and they always start just checking in, which is like the instant.
The instant whiff of desperation of a spurned lover, right?
It’s like, you know, just checking in.
Please, can you talk to me?
And by the way, please buy my stuff.
And I was watching some of the stuff that Vanessa talks about.
And she talks about how this is just an instant turnoff, not just personally, but also in business.
And she talks about some of the solutions she recommends.
So the next time I had to write an email like this, I wrote my email.
I went to chat GPT and I said, this is what I want to do.
And then I said, tell me how Vanessa Van Edwards would write this email.
Very cool.
And it came back to me and said, Vanessa would never write such a weak, weak email like you did.
It would start like this and then it would go here and then it would project warmth and then it would project competence, etc., etc.
And so now, every time all my emails of this nature start with, hey, Mike, Michael, the last time we spoke, we said, or it says, I was thinking about you because the following thing happened.
Right.
And my email reply rate, not just because of these two sentences, but because I’m running them through the filter of Vanessa, has doubled.
That’s awesome.
Wow.
Great.
And you can do this with any expert who is visible on the internet and who has been digested by the all-knowing large language models.
I’m going to run my emails through Larry David.
That’s a brilliant idea.
All right.
So, Andreas, we were talking at the top about ways that AI is being used in the workplace at the moment. chatbots, pattern recognition, so on.
So one of the uses for AI and machine learning that Six Seeds is utilizing, you talked about at the Outlook conference.
So can you talk about the AI personas?
What are they and how do you use them?
Absolutely.
And tell me if I’m blathering on too much because I can talk about this for hours until people fall asleep around me.
So the idea is that you can create personas that are based on real people and then you can feed them the data that we have access to and then you can ask them questions.
And it’s something I showed how it works at the Outlook conference.
So I said earlier on that we can fine tune our data output to people in all kinds of different stages of their life.
If they’re a man, a woman, income, location, political affiliation, the whole thing.
We can sort of say that is important to that kind of person and we can do this with a great degree of certainty.
And then so what we started doing, we started saying, well, here’s a backstory for a person.
And initially, we paid people 200 bucks for two or three hours of their time and we had about 200 questions we asked them and that was really affected.
And to build a backstory for a person.
And then one of our engineers said, we don’t need any of this.
He said, we have the data.
We can build these in a fraction of the time and for no money.
So we started testing that.
And so now we have a Michael and Andreas and a Leslie and we know their background and we know what they like to do.
We know that they’re married.
We know what things are important to them, if they like to drive a car, they like to cycle, what they like to do in their spare time.
So we have these really rounded personas.
And then on top of this, what we do, we take our custom data, our custom food consumption data, and we don’t load everything in there.
So, but for example, for rice, we can say, you know, what’s important to Michael and Leslie when it comes to rice consumption?
What kind of food do they like?
What kind of recipes really speak to them?
How often do they eat rice?
And what, you know, and what kind of rice is important to them?
Is it important for them?
Well, in your job, I’m sure it is.
You know, does it have to be US grown rice, et cetera, et cetera.
And then once you loaded this into your model, you can then interact with the model.
And the first time we did this, and this frequently happens when you work with AI, I had to sit down because the quality of the answer was just so unbelievably high.
And it wasn’t just regurgitating data.
It was speaking with an emotional maturity that is difficult to get out of real people.
Now, I’m not in any way trying to denigrate market researchers because they’re doing a fantastic job.
And very, very often, the value of market research is as much in the question as it is in the answer, right?
But the wonderful thing about the AI models is that they don’t feel the need to perform, right?
They are pulling from the data that we have been feeding them.
They then mix it with the emotional backstory that they have access to.
So, you know, I think the persona I showed in Little Rock was Sam.
Sam’s persona is she’s half Chinese American, half Caucasian.
And she said something about using Asian rice for the recipes her grandmother taught her.
And then she said, and I feel this in my bones.
Right.
Right.
That’s a really powerful statement from an AI, but it shows you how rounded the personas are.
Now, imagine, and we’re getting closer and closer to this, and certainly in my own company, we think of AI as our coworkers already.
Like, we have AI coworkers.
Imagine running a rice company and having six, 10, maybe internationally, you have 50 from all different countries.
People that you can ask questions to.
What’s important to you?
What kind of flavors do you like?
Take a look at this packaging.
How do you react to this?
What’s important to you?
And all of a sudden, and all this is, it’s just interacting with data, but you’re interacting with it through a medium we’re all used to, which is talking to another human being.
Right.
So rather than sitting on a computer and looking at graphs and ripping out your hair because nothing makes sense, you can ask your Michael persona, your Andreas persona, or your Leslie persona, and say, hey, you know, you are, you were born in Europe.
You like rice.
You like to cook.
You have the following flavor preferences.
You have three kids.
Tell me what’s important to you and does this product appeal to you?
And I think, Michael, your kids are a little bit older, right?
Are you, are you an empty nester?
Yes.
Right?
So I probably buy more bulk rice than you do.
Yes.
At this stage in our lives.
And perhaps, perhaps you have the freedom to experiment more with flavors because, let’s say, a recipe doesn’t work out.
It doesn’t matter.
It’s just you and your wife, right?
And you can say, ah, whatever.
But I have three hungry teenagers screaming through the house, right?
And if anybody has ever sort of met three teenage boys, we can’t buy food fast enough these days.
You know, we have, we have just like, I have, I have a, I have a conveyor belt from our local running into my kitchen.
Yeah.
And even, even that’s not fast enough.
But so, you know, those different needs in, in different life stages can be very, very clearly defined with AI.
And then I can just have these conversations.
And as I said earlier on, the more you talk to AI, the more you look at it as a partner in exploration, the better the outcome will be.
What’s your level of confidence with some of these, with these personas?
It’s, it’s funny you should say this.
So, this was always one of our pain points where we were like, you know, this, this feels great.
Yeah.
And then, thankfully, somebody at Princeton did a study about it.
Completely, completely with Albert on it.
And they did a study about it.
They tested 1000 AI personas versus 1000 real personas.
And their overlap was between 85 and 92%.
Well, that’s, that’s great.
I mean, that’s, that’s industry standard.
I mean, I remember, I remember actually a rice, a consumer focus group that we were doing.
We were in, I think Debra and I were in Battle Creek, Michigan, or maybe Betsy was with us too.
Anyway, we were, you know, behind the mirror watching the group go.
And it was, you know, a couple of issues they were talking about.
And we had invited a member who was in the area to come by and watch and, you know, sit and eat M&Ms in the dark room with us behind the glass.
And somebody said something and they were really getting into the point.
And we had the moderator in there and the researcher back with me.
And we’re sitting there and we’re taking notes and we had been doing this, traveling all around and doing this.
And the person said something and the member, our member went, oh, like got excited by it.
And I remember looking over and saying, she’s lying. and everybody, all the research, we all sensed it, that she was lying, that she was performing for somebody else in the group.
And so, we didn’t have to say anything.
The moderator knew to kind of drill down on it and kind of get her to explain her reasoning and when she couldn’t, we knew she was lying.
And we’re like, that’s, but I guess you don’t really have that as much.
Well, absolutely.
And that’s why I say, you know, this is in no way meant to denigrate moderators and people in research who are experts in their field. if anything, it’s again, it’s an additional tool in our toolbox.
Right.
You know, should you run this against a panel, run by an experienced moderator?
Absolutely, you should, just like your recipe should be tested by a real chef.
Right.
But once you’ve satisfied yourself about the quality of the responses, now having access to your focus group 24-7.
Yeah.
And now saying, listen here, product developer person, log in here.
Now, test chef human, log in here.
Now, marketing person, log in here.
Right.
And throw your questions at this.
Yeah.
That is a game changer in information and making it again.
You know, I’m talking about sense making and meaning making.
The meaning making happens with the chef.
It happens with the marketer.
It happens with the new product developer. happens with senior management.
Right.
And now I have an entire panel of humans.
And it’s not the only tool, but this is just one of many tools.
I have this entire panel, this AI panel, that brings the sense to the meaning. and it helps humans to be more human.
I said early on, I have, I have teenage, you know, three teenage boys in the house.
And my recommendation to them will be to study the humanities, study philosophy, study languages, you know, throw in some religious studies, figure out what makes humans human.
Yeah. in all different shapes, sizes, backgrounds, because in the age of AI, that will be a very valuable commodity, very valuable knowledge.
It makes me feel a lot better about my liberal arts degree, so I appreciate that.
I was just reading this morning about something about it, that STEM only isn’t going to do it, especially as we, and we were seeing examples of it in Washington.
All right, we’re getting tight on time, I know, so I want to ask you about another example that you gave that I thought was a great example.
This was something you talked about when we were with the Rice Millers at their convention.
You talked about the air fryers.
Oh, what’s fun, yes.
The air fryer trend.
Can you just tell that story?
Because I think that’s, that’s a great example of how it, kind of, where all this, you know, rubber meets the road.
Yeah, absolutely.
So just to sort of be very quick about it, we had a client from Brazil, and there is a highly culturally relevant product in Brazil.
It’s called, please, if somebody speaks Portuguese, do not write angry emails because they’re going to murder this.
It’s called Paiutiqueo, cheese bread.
And it’s a naturally gluten-free product product that makes it into a little cheese puff.
And in Brazil, it’s part of, like, every child grows up with this.
It’s a snack, and it’s delicious.
I mean, cheese, you know, how bad can it be?
And this particular client of ours had an issue.
So they had a frozen product.
These were like frozen little dough balls.
And they got listed in retailers and they got delisted because they couldn’t hit the velocity.
They couldn’t hit the sales numbers that the retailers demand.
And they came to us and said, we need to run a marketing campaign.
We need to educate people about this, et cetera, et cetera.
And I said, well, can you bring some samples?
And I brought some samples.
I chucked them in the oven at home.
And the instructions were a bit of a kerfuffle because you had to heat up the oven and you had to preheat it and then 425 degrees and, I don’t know, 40 minutes, yada, yada, yada.
Take them out of the oven.
They were absolutely delicious.
And then I thought to myself, because I am a culinary geek, I thought to myself, I wonder what happens when I chuck them in the air fryer.
And I threw a handful in the air fryer.
I air fried them for a couple of minutes and they came out.
And to me, they were indistinguishable in quality to what has just come out of the oven but in 10 minutes rather than 45 minutes.
So I called up the clients and I said, listen, have you ever tested those in the air fryer?
And he said, no, I have not.
And I said, could you do me a favor?
Could you go out?
Can you buy five air fryers?
And can you buy a cheap one, a really expensive one and three in the middle?
And he said, sure.
And I said, I’ll come by tomorrow.
And so he bought all these air fryers, put them in the factory, we tested them, absolutely delicious, not a problem.
And so we then sort of looked at, we used our data and AI to figure out the pain points that people associate with products that they have to use an oven for.
And those were, our kids can’t use it by themselves.
Kids are not allowed to use the oven by themselves. it feels like cooking.
It’s expensive from an energy usage, which was interesting.
I had never seen that before.
And what was the other one?
Kids, energy, time.
Oh, and it takes a long time.
Yeah, time, yeah.
Yeah, and it was takes a long time.
And so rather than running an ad campaign that would have cost them an awful lot of time, they were at the, they needed to print new packaging.
And we said, look, why don’t we just put air fryer instructions at the back of the packaging that addresses these pain points?
And why don’t we talk to our retail partners and we run a program with them where we just slap some stickers at the freezer door as a frozen product.
Within six months, their velocity had increased by 18%, 1, 8% through those very, very simple changes.
And so that’s just an example of how understanding consumer needs, not reported consumer needs, but behavior needs, can change the fortunes of a company quite significantly. you know, at a cost that’s less than the cost of marketing.
There is some Nielsen data I saw last year where Nielsen reported that 70% of all marketing campaigns in CPG and food and drink cost more than they bring back in money, than they bring back.
That’s significant, especially for smaller brands.
If you go out and you listen to your consumers and you figure out what really bugs them and you then find a clever way to deliver against that, that’s a really good way to run your business more effectively.
Yeah, that’s awesome.
All right, we’re going to take one quick break and when we come back, Andreas is going to look around corners and predict the future even more than he already has.
This is Deborah Willenborg, Deputy Editor of the USA Rice Daily Newsletter.
The fact that you’re hearing this means you are interested in rice.
And if you’re interested in rice, you should also be reading the USA Rice Daily.
Every weekday at 2.30 Central and 12.30 Pacific, our newsletter publishes excellent reporting on news that’s important to the rice industry.
From the latest trade deals to promotion programs, food trends, and from regulatory action to sustainability stories and industry events.
If it matters to rice, it’s in the daily.
We also include a roundup of ag news from other sources.
Rough rice prices updated every day, a market year average price tracker updated monthly, and other relevant information such as links to the USA Rice COVID-19 resource webpage.
And of course, there’s our quote of the day to put a smile on your face or inspire you.
So make the USA Rice Daily part of your day.
It’s easy to sign up.
Just go to usarice.com and click the box that says more news to receive a daily dish of rice news.
All right.
All right.
Welcome back.
We are talking with Andreas Deuce, CEO of Six Seeds Consulting and a frequent USA Rice event speaker.
Andreas, you were talking before about the virtual panelists.
It’s not something that I imagined five or six years ago when I was sitting behind those one-way mirrors like I was talking about.
What is a new tool or development that you think you’re going to see or you’re hoping to see in the next like 18 to 24 months?
Oh, my God.
I think what we’re currently seeing in AI is the equivalent to what happened with early Steam engines.
And I don’t have the name in my head.
It’s called Something’s Paradox.
Maybe you can quickly Google it or something.
It’s Something’s Paradox.
So what happened with early Steam engines?
When Steam engines first appeared on the market were a thing.
The only place you could realistically run a Steam engine was at coal mines because they were so expensive to run and they needed an enormous amount of coal to run anything.
And then James Watt came.
And so coal mines made a lot of money and there was like there was, you know, the shares for coal mines were high. and then James Watt arrived and he made some changes that dramatically increased the efficiency of steam engines.
He put insulating around the boiler.
He did a whole bunch of things.
So all of a sudden coal consumption for steam engines dropped by 75% within three years.
Like that was really, really, really fast.
And shares in coal mines crashed because people said, well, now we don’t need more coal.
But what happened was the exact opposite.
Steam engine use absolutely exploded because all of a sudden it was practical to run steam engines away from coal mines.
So now you could build trains.
Now you could build agriculture like the first tractors were steam powered, right?
You could put power stations in cities.
You could have small factories and farms using steam power for the first time.
So steam use use of steam absolutely exploded and so did the use of coal.
Now a lot of those, when you look at the history of inventions, a lot of the inventions that were made at the time are absolutely ludicrous today.
Steam powered watches and all that kind of stuff.
You know, it was exactly what’s happening today.
Everybody is slapping AI into something, hoping to make a quick back out of it.
And engineers in the 1800s did exactly the same thing except they slapped a steam engine into everything that could possibly, you know, accommodate one.
So what we’re going to see in AI, this is my prediction and I’m fairly certain that this is true, that AI as a thing will almost disappear.
When I walk into a room and I switch on the lights, I don’t think about electricity.
I don’t get all excited about my electricity provider.
When I switch on my computer so we can have this conversation, I don’t say, gee, gee, my power supply, that’s a great company doing interesting things.
All I’m thinking is, but I want to talk to Leslie and Michael today and have a fun conversation about AI.
And the same thing is going to happen with AI, that people will get less excited about what AI can do or what the AI bit and they get excited about the meaning making.
Where can it take them as human beings and how can it support what they really want to do?
I’ll give you an example.
My oldest son is quite heavily dyslexic.
So words dance in front of his eyes.
He’s a terribly bright kid, but reading and writing will always be a problem for him for as long as he lives.
With agreement from his teachers, we take his school Google slides and we upload to a tool called Google Notebook LM.
I don’t know if you’ve heard of it, but one of the things that Notebook LM does, you can put anything you like into it and it creates a completely artificial podcast from the content you put in.
You cannot tell that these are synthetic voices.
I will absolutely challenge anybody to say, tell me that this is not a real person talking.
And it takes this content and it turns it into a conversation between two people.
And so for him, this is an example where we don’t get excited about the AI, but I’m really excited about my kid now being able to learn in a way that supports him and makes the most of his abilities.
See, it is coming for our jobs, Michael.
It’s going to replace us.
Yeah, okay.
You know what?
Yes, I mean, AI is coming for jobs, 100%.
But having said this, if you look at the census from whenever the last census was in the US, 65% of the jobs on those census didn’t exist in 1965.
Right?
Right?
so old jobs will disappear just when Henry Ford started putting the Model T on the road, horse-related jobs disappeared.
When the printing press entered the market, there were riots in Paris.
There’s always a riot in Paris, but there were significant riots in Paris.
And people were burning printing presses on the street because the scribes were like, well, what are we going to do to feed our families?
Right?
So these things, I don’t mean to make light out of personal hardship, you know, and these things, yes, jobs will go away, but other jobs that we don’t even know about today, they will appear.
And the trick really is to look at this as a new kind of electricity, a new kind of empowerment, and then don’t get excited about the electricity bit, but get excited about the stuff I can build with it.
Yeah.
That’s here.
End us the sermon.
Well, I think that, I think this has been an amazing conversation.
I have got, like, a lot of this is going to be floating around in my head for days, I think, kind of as I turn it over.
So, as chat GPT predicted, this has been a, been a great conversation, great episode.
Thank you so much.
Yeah.
So, Andreas, thank you so much for coming on with us.
Where can people find more from you and Succeeds?
Do you know what?
Find me on LinkedIn, my messages are always open.
So, ping me on LinkedIn.
I have an unusual name and I’m easy to find.
Also, you can Google me.
I Googled myself recently.
I don’t think there’s anything embarrassing that will turn up.
I’m old enough that all my useful transgressions never made it on the internet.
Right.
So, they haven’t digitized the pictures or anything.
They haven’t digitized the competitive wine drinking by the lake house.
Right.
I’m very glad.
Right.
Right.
Right.
Well, we appreciate you coming on here and giving us the full time, not a 25% subtraction of time for any tariffs that might be because we know you’re coming to us from Toronto.
But thank you, Andreas.
Thank you for coming on The Rice Stuff and thank you for listening to The Rice Stuff for USA Rice.
I’m Michael Klein.
And I’m Leslie Dixon.
We are online at thericestuffpodcast.com.
You can email us at podcast at usarrice.com.
We also hope that you will subscribe and tell folks about us.
The Rice Stuff is produced by USA Rice and Human Factor and was engineered by Blake All.
.