How Amazon and Walmart may be disrupted by ChatGPT and Perplexity

Now, it is on another bet on the e -commerce for the era of artificial intelligence after spending most of the past decades in launching and operating the car service work on demand. His new company, Refibuy (an abbreviation of research, research, and purchase) aims to create programs that help brands for consumers and retail dealers to move in a new world via the Internet where shopping agents of artificial intelligence make their purchases as human customers do.

luck It was recently discovered with Wingo to discuss how artificial intelligence agents and AI are disabled online shopping, and how this moment can be affected in e -commerce. Not surprisingly, as a person trying to build new new business in space, wingo may be more polished than many about how new new shopping tools may disrupt the artificial intelligence giants of e -commerce and old retailers alike. The interview was edited for length and clarity.

luck: What did you see in the e -commerce that made you want to return to?

Scott Windo: LLM companies [like ChatGPT, or Perplexity] The consumer’s attention has the attention of the consumer, so they get the distribution of the consumer. And when you get this consumer distribution, you have to discover, how can I invest it? It has become very clear to me – especially when I launched the first confusion [shopping feature] -This will be the next large wave of e -commerce, and I could not miss it. I feel as if I have a unique view of the subject and see how electronic commerce has been done so far-good and bad.

What is the “bad” that you refer to?

If you look at e -commerce in the United States, the level of growth may really slow down, right? Based on the numbers you look at and what you calculate, we settled somewhere between 15 % and 20 % of the total retail sales of e -commerce, and this was a very fixed number.

But if you look at other countries, this number feels wrong. You look at Europe and this number is higher. Asia is the highest way than the charts. So why is this? I think this is because our user experiences have really been charged or stagnant. You are a great follower of Amazon, as I am, but something happened as they completely stopped creating a retail and just began his critical criticism.

Since Amazon mentioned, how do you see that she is affected by this new world?

So they have a retail trade, a [third-party seller] market, ads.

The advertising layer in my mind does not add value to the consumer experience. So imagine a world in which someone builds [an AI experience] In addition to Amazon and no longer uses Amazon Front Write – they started a chat or agent of Amnesty International. Amazon just the back interface now, so humans do not see product ads on Amazon. So there are 60 billion dollars from the profit margin mainly. This is the problem number 1.

Problem No. 2 is now you are just the back interface of a different front facade [like ChatGPT or Perplexity]And at some point you have to pay a tax [to the AI company]Because the switch can be done. So the Amazon hates paying a tax. They are the tax collector, right? So there will be a battle for those who are the tax collector. So this will be interesting.

Then I think the third thing is, if you are ChatGPT or [OpenAI CEO] Sam Altman, what if you only collected a 5 % price for sellers to include their products (compared to 12 % to 15 % for many categories on Amazon)? “You can keep your products listed on Amazon, but we also love them to be included with us as well. And since we are cheaper, are you thinking about offering a lower price?” A brand like Dyson [might] Do not take this deal, but the small businessman may take this offer.

So there is a way in which you can attack Amazon by quickly getting rid of advertising work, which will be largely harmful. Then I think you can chew in a large part of the market business. And those are the two – AWS aside – the company’s profit centers.

When you talk about someone based on Amazon, you are talking about search engines of artificial intelligence or shopping agents who become a starting point for more customers to shop online. How will you search for products or shop differently through shopping agents of artificial intelligence more than we do today?

First, the multimedia will be – it can be a sound, write it or anything you want. I think we will be attracted to the expression because it will become really good.

You can do this conversation with what your personal assistant will be, and you will do everything for you. So, you will have a conversation with her about your calendar, and she will tell you, “Remember that you got this dinner tonight, Bob will sit to your left, which is CTO for this company and remember that you have this story out of his company.” You will basically become an advanced assistant.

Then he will say, “By the way, can I have a quick conversation with you about the things you need to renew?” You will say, “definitely.” It is like, “Well, you got the toilet paper on this day, milk on this day, do you agree?” You will say, “definitely.” Or you will trust him a lot, it will be requested only. This is how the basics will happen. It will be just a quick conversation or you set it on the automated pilot.

And in the back end, who is participating there?

Perhaps he is Amazon. Perhaps Wall Mart. The agent or assistant will decide that for you. You will give her parameters, perhaps on value or comfort. So the value will be, “I don’t care from where the milk comes from, as long as it is under X Bucks Gallon and 2 %.”

Then it becomes really strange, because now the agents can run small auctions and talk to each other and say: “I got this big arrangement, who wants it?”

On the comfortable side, you buy something new, which is a deliberate purchase – for example, you want to spend $ 150 on the paddle of the pickled ball. You will know the right questions to ask and suggest the features that you need to find. Perhaps you like it, “I need it in two days, and I need to pay it [the Buy Now, Pay Later option] to be sure “.

How can the new Amazon features of artificial intelligence like Ruffus AI help their battle?

It’s good efforts, but I think Chatgpt will be better. Part of it is that Rovos cannot be so good that you do not need [Amazon’s product search] propaganda. In order for Rufus to improve significantly, you must replace the current search experience on Amazon. This will not cross Amazon, as this would kill $ 60 billion of advertising revenues that are primarily a pure profit margin.

Do you have any meaning that you are doing that Amazon or Walmart is aware of the threat, at least as you see?

I haven’t had conversations with Amazon, but Walmart people have publicly said that we will take a very open approach to shopping agents and welcome them. Somehow, this is smart because they are a competitor to Amazon, and I do not think they have anything they lose. This is an opportunity for them to settle the play field and a kind of catching Amazon. They are fine if the front end of all these transactions is not, because they also have stores and many of their orders are taken online in the store, so at least they will have the consumer interaction in that layer.

What is your message or two retailers or brands about how to change their operations, or what they should do in a different way from their action now?

So the sincere answer is that we believe this is the beginning, and it will be bad, and what you have to do is embrace these agents, and your data is not ready for them.

There is a set of data we have in the retail world, all of which are not really ready for agents. So there is no criterion for product. This is what I lived in the channel “The Channel” – everyone will send us the catalog of their products, and they were hot chaos.

Choose any category and everyone contains different features (properties or details that describe a piece of goods) for each product, and they use different terms. Let’s say that the gold standard for GoPro 10 had 200 cm. Any certain division merchant may only have 20 of these. Perhaps if you are Amazon and Walmart, you have 50 of these features.

But the agent engines need more content and context. The average search for the product may include four terms, but artificial intelligence inquiries or chats are 15 to 20 words. To bridge these things, we want to help retailers look at their data, evaluate them, and a kind of saying, here is the place where you are fine, and here not you, here you are ready, and here you were not very ready. Then we will have capabilities to go and help.

How quickly all this happens in your point of view?

I started thinking about this idea and talking to people in the fourth quarter of 2024, but over the past thirty days, it has ended from a truly large retail type of billions of dollars in retail traders to “I am receiving pressure from all aspects to find out.” So there is a pressure level pressure. There is a lot of that. But what they really see is that the distribution of incoming traffic changes very significantly and very quickly. In May, it started in e -commerce, people behind the scenes, and hair on fire, and they are afraid of that.