In this podcast, Motley Fool senior analyst Jason Moser discusses:

  • Shares of AmEx popping 15% in two days off of strong guidance for 2023.
  • How younger customers are the majority of the company's growth story.
  • Why Amazon's revenue number is so important.

Motley Fool engineering manager Tim White and Motley Fool senior analyst Tim Beyers discuss the implications of Microsoft's $10 billion investment in ChatGPT.

To catch full episodes of all The Motley Fool's free podcasts, check out our podcast center. To get started investing, check out our quick-start guide to investing in stocks. A full transcript follows the video.

This video was recorded on Jan. 30, 2023. 

Chris Hill: It's time to dig into that chatbot everyone's talking about. Motley Fool Money starts now. I'm Chris Hill. Joining me in our brand new studio it's Motley Fool Senior Analyst Jason Moser. Good to be here with you.

Jason Moser: It's a beautiful day.

Chris Hill: It is a beautiful day. It is a down day for the market at the moment, but shares of American Express continue to rise. Fourth-quarter profits and revenue were a little bit lower than analysts were expecting but the guidance that Amex put out for the new fiscal year was pretty strong. Between Friday, which is when we got these results and today, shares of American Express up 15%. This is the best run in the short term that this company has had in a while.

Jason Moser: You don't see a business like this make that move in such a short period on typically. But I do think the theme here for the quarter, for the year, the US consumer hung in there. For the quarter, for the year versus the other sides of the business, the commercial services side, the international side. There were some currency effects, of course, that played in the international side of the business, but they just continue to build itself into a brand that resonates beyond just the lifestyles of the rich and famous, which is what they were known for so long. I'm going to save you my Robin Leach impression and the listeners as well and let's just move forward. Looking at the numbers. I think they were very encouraging.

Revenue for the quarter was up 17% to $14.2 billion. That was driven by consumer spending, primarily earnings per share of $2.07 compares to $2.18 from a year ago. They saw a little pressure on the expense side. We'll get to that just a sec. Generally speaking, this is a very similar narrative to a lot of the banks we've been talking about. Remember, American Express is a bank holding company, so they are beholden to those rules and regulations. Net interest income up 31% for the quarter. Not terribly surprising given the rate environment but also remember American Express benefits from primarily fee-based products, those fees that come with owning those cards.

But network volume for the quarter up 12% was $413 billion. We talked a lot with the big banks about provisions. They're seeing some losses coming down the pike here and so consolidated provisions for those losses for American Express, they were $1 billion compared to $53 million from a year ago. Not terribly surprising to see there. If you look at the total reserves, total reserves ended the year at $4 billion, that's about 2.4% of total loans and card member receivables. That's essentially flat with a year ago on a percentage basis. Again, back to the expenses, those were up 15%, saw a big boost in travel-related benefits, which I think makes a lot of sense. Folks are getting back out. But all things considered, I think this was a really respectable quarter.

Chris Hill: It was. Amex keeps its streak as a dividend aristocrat intact. They also announced a boost to their dividend at 15%. It's interesting to see, is it an overstatement to say a demographic shift? We've talked in the past about Amex in some ways, being an older brand attracting older customers and you dig into the result, it seems like they are doing an effective job of appealing to younger people.

Jason Moser: They really are. I think that was the big question mark for a long time as it would they be able to do that in this new FinTwit-driven environment where they're seemingly a brand new offering every day. They've done a tremendous job. They acquired three million new card members for the quarter but if you look at the full year, new card acquisitions grew 12.5 million members and nearly 70% of those new accounts acquired are on their fee-based products and you know how powerful those fees can be.

But when you look at attracting that younger demographic, I was really impressed to see this, millennial and Gen Z customers are the largest drivers of growth for the business. They represented over 60% of consumer card acquisitions in the quarter and for the full year. That was a big question mark we had with the Amex several years back. Fast-forward to today it really does look like they've built this suite of offerings to really serve everyone out there. That really matters in this day and age when you think about how much competition is out there, how many offerings exist today? American Express is still holding a strong position.

Chris Hill: Particularly in the wake of the most recent wave of big bank earnings and part of the narrative, there was the struggles that Goldman tax and others are having with their more consumer finance-based tech solutions that they were shelling out tens of millions, if not hundreds of millions of dollars for.

Jason Moser: That's a good point there. When you look at something like Marcus clearly something that was built with the younger consumer in mind. Clearly also not gaining very much traction. Goldman having really pulled back on all of those investments. American Express, I think for them they have a bigger playing field. I think they can experiment more. It really I think goes to show you too the strength in the cardholder rewards programs. You can pull different levers there and try new things which clearly they're doing.

You get back to the stock, you mentioned that dividend, I think is terrific. Also, they continue to repurchase shares. Share counts down about 13 1/2% since 2017. When you look at the stock today, given the guidance for the coming year at the midpoint that puts shares around 15 times full-year earnings estimates, it's not terribly outrageous given the quality of this business. You'll look at the last year the stock has held flat, but you go out three and five years, it starts to look a lot better. Ultimately five-year returns stock's up almost 90% and better than doubling the market. Things have been working out well for patient shareholders here.

Chris Hill: Twenty-one percent of the companies in the S&P 500 are reporting their latest earnings results this week. Apple, Alphabet, Amazon, AMD, Amgen, that's just the ace. What is one company you're going to be watching this week maybe a little bit more closely than others and what are you going to be watching for or listening for on the call?

Jason Moser: Amazon's always an interesting business because there's so many facets to it. Amazon to me, when earnings come out Thursday after the market closes. I think it's going to be fun to watch how the market reacts and really want to pay attention to what management's saying here. If you look at last quarter, they saw sales grow 15% and that growth was essentially flat with the previous year. Now, they're guiding for revenue in the range of 2-8% this time around, which would compare to growth of 9% from the same quarter a year ago. But when you look at the challenges the business is dealing with today, you get the retail performance that I think is raising some questions, probably more questions than something like AWS. AWS continues to succeed. It's the lion's share. It really is all of the company's operating profit, to be honest.

Jason Moser: I think looking at the retail operations for Amazon; a state of the retail operations today, they had a glut of warehouse space. They've had to figure out how to right-size the business in some ways. Obviously, they are undergoing some layoffs as well. I think getting an idea of the state of the retail operations. You had a lot of talk here these days about what if Bezos decides to maybe come back and try to help this business get fined its bearings? I think it's an interesting conversation to have. I'd be very surprised if it happened though.

It's totally possible, but you have to ask yourself the question what would Bezos do differently than Jassy is doing right now? A lot of what Jassy is cleaning up occurred still under Jeff Bezos. I'm not knocking Bezos, a lot of companies made a lot of decisions based on some abnormal circumstances. Jassy's in the middle of cleaning that up, so I'd be surprised if that narrative of perhaps Jeff Bezos comes back. If this is a dot of a quarter, I don't know that's really the solution but getting a better idea on the state of the retail operations there are they making progress and rightsizing the business? It's going to be a very interesting call for sure.

Chris Hill: Do you think he wants to? He seems like he's having fun with the outer space.

Jason Moser: I just cannot imagine you want to get back to that day and day.

Chris Hill: Jason Moser, thanks for being here.

Jason Moser: Thank you.

Chris Hill: Chatbots have been around for years without a lot of fanfare, so why is Microsoft investing $10 billion in OpenAI; the creator of ChatGPT. Tim Beyers and Tim White look at the implications of the deal and how big a threat it could be to Google Search.

Jason Moser: Tim we've got to talk about OpenAI which is the company behind ChatGPT which seems to break news every single week. This week we are talking about Microsoft's and I should say, reported not confirmed, $10 billion investment in OpenAI to take a fairly big steak and presumably have some influence over ChatGPT. What is your instant reaction to Microsoft stamping its foot in this space saying, look, Azure is going to be the cloud for ChatGPT and that's the way it is because we just wrote a $10 billion check.

Tim Beyers: I think Microsoft really had no choice. They believe, like a lot of people to do that generative AI, like ChatGPT is capable of competing with Google in terms of people's attention in getting the answers to questions. It already is winning hearts and minds among programmers in particular. To get answers to coding questions into even write code for them and I experimented with this quite a bit, and it is pretty good at writing basic code from a relatively small textual description. Essentially what Microsoft wants to do is bake the idea of generating text into every one of its products. You can imagine you're writing and outlook email, and you can say something like draft a memo to this person that says basically this and it rights out the text of that for you and actually rights all the pros and does that right within Microsoft outlook.

Jason Moser: Let's talk a little bit more about how ChatGPT functions like we call it a generative AI. Let's, define that and let's define ChatGPT and why we think this has such a big lead here in this space.

Tim Beyers: ChatGPT is an interface. It's a way of interacting with this AI as a chat bot. You write it some text and it right back to you. Like a chat stream, like you're texting back and forth with it. ChatGPT stands for chat obviously, and then GPT is generative, pre-trained transformer. Pre-trained, I think, is a really important part here, so pre-trained means that this AI model has been trained to be able to answer general knowledge questions. OpenAI is goal is to make general knowledge AI. To do that they have to feed it a whole bunch of information. They stopped treating it information at the end of 2021. It really doesn't have any current events type information.

They can't answer like, what is the weather today or anything like that. It can only do general knowledge from pre 2022. But for things like what is the right way to program the software that really hasn't changed in a few years so that it's really easy for it to do. But the whole idea of generative AIs, it's generating something. It's creating something. In the summer of 2022, we had the explosion of image generative AIs such as Dolly, which is also from OpenAI, that creates images from texts, descriptions and which had GPT does is create text and not just text, but like a whole essay or hold description of something or whatever you want it to create from a chat prompt.

Jason Moser: I'm going to draw a comparison here. Then we should talk a little bit about some history because as you and I have talked about AI on the show in the past, there's the race in AI, it's a little bit of a boat race in the sense of the AIs, whether or not they have the most elegant design or not. The AIs that get used the most are the ones that are likely to win because they will have the most data and the more data they have, and the ability to draw conclusions and have a library of conclusions based on that data will fundamentally just make them better.

They will become better AIs because of it. This ChatGPT in some ways, it's a bit like, and I think you've made this point before, 10 like Google in 1999. What is this thing? Of course, Google has gotten better because we've used it for close to a quarter century, for trillions upon trillions of searches. In a lot of ways, what OpenAI is doing by making ChatGPT free and openly available is essentially saying, here's my toddler, he knows some things. He's helpful to you in some ways. But if you could train him up and ask them a lot of questions and tell us when he gets things wrong and that sort of thing. He'll be a big boy someday. That is what they're doing is they're basically opening them to the world to be trained by the whole world.

Tim Beyers: You are talking about. You are basically likening ChatGPT to the little kid that will not stop asking why.

Jason Moser: Yes. The precocious child is what ChatGPT is and OpenAI is saying, please train him up for us, or, train them up for us, and have them be smarter every day. That I think is really important to realize that people ask, why are they making this free open available? Because just like every time you search, Google figures out which linked you clicked on and makes that link a little higher and the results next time. Every time you ask it a question, they're making a note of that essentially in saying like we need to make sure that ChatGPT can answer this question accurately.

Chris Hill: It's really interesting and we made the Google link obviously for deliberate reasons because some of the history here, and I do want you to talk about this and we could talk about some of the backing here. I think the most interesting part of the story is the founders. But another interesting just financial piece of this story if this company does go public at some point and we're talking now about OpenAI, Crunchbase reports $11 billion in private equity funding so I would expect the valuation is not cheap. Ten billion of that from Microsoft, but a lot of very well-heeled investors. Sequoia Capital, Andreessen Horowitz, Tiger Global, Khosla Ventures, Y Combinator, Reid Hoffman, one of the co-founders of LinkedIn. There are some really big investors in this, but I think part of that has to do with some of the history here and the founders, let's talk a little bit about that because it's got a Google connection.

Tim Beyers: Eliya Sutscava is a Russian software developer who has been in the United States for quite a while now. He worked on a bunch of projects at various university things that eventually got bought by Google in the early 2000s and turned into what they called Google Brain, which was their internal AI. Then they also bought Alpha, which is known for the DeepMind products. They combined that all into their AI world.

Essentially, Eliya was one of the authors on the AlphaGo paper, which was the huge like we made an AI that can win against the world champion at Go, which is a relatively difficult game to program things to win at. As a result, they had a lot of really high-powered AI folks. Well, Eliya left along with a number of other Google Brain folks when Alpha became the winner of the AI world at Google and went off along with Greg Brochmann who came out of MIT in 2010, only 13 years ago and created OpenAI and one of the co-founders is also Elon Musk. Essentially a big part of Google's AI brain trust, literally their Google Brain Trust came out to found this and now people are making noise like ChatGPT is going to be the Google killer.

Chris Hill: Let's talk about that. Let's talk about implications here because we've had this discussion on the show. I'll let you go further on this because you've made this point more elegantly than I have. But there is this point at which Google gives you the way to search for yourself what you want. Then if you have AI and ChatGPT, just doing things for you and serving you. So I don't even have to ask, you just eliminate the ask from the entire process. Is that what makes this so threatening to Google?

Tim Beyers: Yeah, I think so. If you imagine that many Google searches are things like, who is this person? What is this thing? How is this code thing supposed to work? If you're a programmer, you spent a lot of time Googling for things. Rather than getting a back-end list of webpages which have all been SEO optimized within an inch of their life. There's also just pages and pages of ads now in Google. You just get some tax that comes back and gives you the answer. You don't have to fair it among all these web pages and figure out, close all the pop-ups and everything to get your answer right. You just get some text that is the answer. This was the promise of Wolfram Alpha many years ago. But ChatGPT goes way beyond the yes or no black or white questions that Wolfram Alpha could do and actually generates not only the answer to a question, but actual text based on the prompt.

Chris Hill: So let's close on this implications for Microsoft here, because Microsoft's Azure, which is its Cloud, will be the Cloud for ChatGPT, which means there is going to be a lot of AI data going forward that is going to exist inside of Microsoft Azure. I wonder Tim, just how big of an advantage this is going to be for let's say the Office products or really every Microsoft product. Where do we see this going?

Tim Beyers: Yeah, I think it could be huge advantage that said, Google has this technology. They have the ability to do this thing. They have just not un-leased it because they're afraid of being wrong. If you've ever seen the people also ask and all the other suggest an answers you see in Google, you'll often those that those are often wrong as are things that ChatGPT will tell you. It's not always right. In fact, it's often not great. The challenge here is that AI isn't always right and won't be for a long time until it gets a lot more training, and that's why OpenAI is doing this. But Google has this technology. They're just going to have to figure out how to get it out in front of people in a way that doesn't make them look like they're always getting things wrong.

Chris Hill: Get ready. Maybe a way to summarize this is get ready for the AI training more's. Maybe Google releases DeepMind, Microsoft with ChatGPT is gathering data and we will be on the verge of gathering more data than we ever have. Microsoft and Google and maybe some others getting into this to gather as much data as humanly possible to understand the world at a scale that we've never tried to before.

Tim Beyers: OpenAI has hundreds and hundreds of job ads open in Europe for programmers and other folks to come in and help them train ChatGPT to be a great coder and to be great at answering questions. They are not just relying on things they find on the Internet, they are actually hiring people to make ChatGPT smarter.

Chris Hill: Well, this is going to be an interesting space to watch. Pay attention we'll be talking about this a lot more. This weekend tack is every Friday from 1-2 PM Mountain Time. Tim and I are in the mountains 03:00-04:00 PM Eastern. Most weeks will be on talking about OpenAI and all other tech topics. We'll see you then. Thanks for tuning in Fools. We'll see you soon. Fool on. As always, people on the program may have interest in the stocks they talk about and the Motley Fool may have formal recommendations for or against, so don't buy or sell stocks based solely on what you hear. I'm Chris Hill. Thanks for listening. We'll see you tomorrow.