Key insights
- The discussion highlights the early stages of AI adoption in financial research and workflow automation. While AI is being used for tasks like quick research, Excel automation, and basic coding, its impact on US equities is currently limited. The focus is on improving efficiency rather than driving investment decisions, suggesting a gradual integration of AI into the financial sector.
AI this and AI that — everyone seems to be using these days. So what tasks are people mitigating to chatbot services these days and how exactly are they using AI?
Yahoo Finance Head of News Myles Udland sits down with Barron's Investor Circle Newsletter editor Josh Schafer, Interactive Brokers chief strategist Steve Sosnick, and StockBrokers.com director of investor research Jessica Inskip for a chat about how they are applying artificial intelligence in their every day lives and work days.
I'm just curious here. Uh, Josh, I'll start with you. Like, where are you at with uh AI and your workflows? And if I can ask, do people say from on high, everyone should be implementing AI in their workflow?
It's definitely something that comes up in the conversation for sure. And I one way I use it mainly as a research tool still at this point. I mean, we have I'm on a stock picking team at the Baron's investor circle, right? When I'm sitting in the meetings, I will just feed it questions that I have about whatever stock we're talking about. What's the bare case here? What is AI actually doing for this company? And just like, it gives you a better answer than Google would, especially in meetings to just give you a couple things to pull out to then be able to ask questions quicker, be a little smarter.
I use it for things that I'm not as good at too. I'm not that good at Excel. So I use, I mean, we're Gemini company, so I'm using Google sheets, but it will help me reorganize the spreadsheet quicker, those types of things. So just kind of basic efficiencies.
From a corporate level, AI is something we we definitely want people using. We we have it, we have it sort of built into our system in a way. I'm not I'm not not 100% sure which model we use actually. Um, because we do a lot of coding. at our root, we are, you know, we're as electronic, we're a software company. And, you know, so we're a lot of our a lot of our programmers are using um AI tools to do the basic code and then then work on top of it.
From my personal point of view, I find AI helpful, um, sometimes to research stuff like that's a little esoteric, maybe it has the answer a little quicker than I'm going to get it or a lot quicker than I'm going to get it. So I do use it there.
The one thing I use it for every day is I write, I try to write something every day, put it on a website, ibkcampus.com.
Everyone in business media very much appreciates your output. It's super helpful.
Thank you. Good. Um and so one of the things I realized was, let me just drop I I never let it write for me. I I haven't tried the experiment of letting it write for me. and I also wish it wouldn't just like complementaly complement me gratuitously, like wonderful piece. But what it does do is it is the best proof reader I've ever found.
Right now, my preferred model is Claude. I have moved over and utilize Claude and I'm in the process of creating skills. So I've already taken my entire technical analysis process and a lot of my market thoughts, the way that I um think about things, even when I write content and I'm creating skills that I put within Claude. So that's step number one.
So I can have it access those skills at any time, which will maximize token usage.
Number two, and it's a theme that is emerging is MCPs. So these are servers where you can connect data. Data is so important for artificial intelligence and we want the right data. We always have a a saying in the tech world, garbage in, garbage out.
So, even brokerage firms are doing this with Morningstar. You have a Morningstar account, uh Factset, uh the Trade station started this. Alpaca has now very interesting. I use Alpaca because that one's a a free server and you can even connect to Yahoo Finance as well.
But essentially, you can create these connections where you can pull in market data and then create an interface on anything that you want. So I actually have three separate interfaces that I use.
One, I have my technical analysis view. I have it this uh I had it create a score of all of the S&P 500 stocks. So it will score them from 1 to 10 and if it's on a the higher the number, the better the buy. And so that has been working great. So I have it with just starting with the S&P 500.
It shows me my technical view on that. I do it by sector, then subsector, just so I can see what's happening under the hood. and I use different moving averages than the normal, so it's difficult for me to do that on other platforms. So I've utilized Claude to do that.