How I Invest with David Weisburd
00:23:54 · Chaz Englander, David Weisburd
So I really think we do an incredible job at working closely during this free discovery phase, whether that's GPs, whoever it might be, identify areas that they should immediately be thinking about applying artificial intelligence. That's what we've become very good at. Now, does that actually look like in terms of use cases? Well, for now, it's more of those inefficient areas of the business that you can have these productivity gains. As I said, I think in future that will become more how you're able to drive insight that wasn't possible pre AI, but for today, it's productivity. [00:24:25] S0: What are we talking about budget wise? What size of firm do you need to be to to hire a firm like Model ML? Or maybe how much do these projects cost? We as a business [00:24:37] S1: don't do many kind of like five, 10, 20 seater type deals. We're really looking at, you know, in the hundreds of seats at the minimum, if not the thousands, right? That's changing and quickly as we're scaling the team is enabling us to work on those smaller deals. And really that's just a question of focus. Again, we wanna make sure we over deliver when we're working with our customers, at the moment it's in the hundreds of seats, I think that's gonna come down to the five and tens within the next few months. Pricing wise, depending on whether you're buying data through us, etcetera, you should really be budgeting for true AI tools, you know, anything from a $100 a month to maybe $300 a month, you know, per seat. But again, it just depends on scale.
00:24:37 · Chaz Englander, David Weisburd
[00:24:37] S1: don't do many kind of like five, 10, 20 seater type deals. We're really looking at, you know, in the hundreds of seats at the minimum, if not the thousands, right? That's changing and quickly as we're scaling the team is enabling us to work on those smaller deals. And really that's just a question of focus. Again, we wanna make sure we over deliver when we're working with our customers, at the moment it's in the hundreds of seats, I think that's gonna come down to the five and tens within the next few months. Pricing wise, depending on whether you're buying data through us, etcetera, you should really be budgeting for true AI tools, you know, anything from a $100 a month to maybe $300 a month, you know, per seat. But again, it just depends on scale. [00:25:17] S0: Chaz, I wanted to do a live discovery. So we have a media side of our business. We have an asset management side of the business. So what what questions would you typically ask a client? And maybe we could role play a bit. [00:25:27] S1: I always ask, this is a question that I think is important is I ask people about their own AI journey as as a as an individual. Right? You know, how... What are they using AI for inside and outside of work? Right? And the reason I ask that is I like to understand their appetite towards AI in general.
00:25:27 · Chaz Englander, David Weisburd
[00:25:27] S1: I always ask, this is a question that I think is important is I ask people about their own AI journey as as a as an individual. Right? You know, how... What are they using AI for inside and outside of work? Right? And the reason I ask that is I like to understand their appetite towards AI in general. [00:25:42] S0: I do kind of look like AI as a person. Sometimes they disappoint... He or she disappoints me and sometimes it comes through. I found... So we have on the media side of our business, we have, a lot of our podcasting, the production, the editing. But we also have YouTube and thumbnail and packaging and all this. And I have found very specific cases where AI is very good, in certain specific cases where it's extremely bad. And even worse than it being extremely bad, it's extremely confidence in those cases. And whenever I ask it how confident it is, it gives me like full confidence and then it's completely off. So that's been kind of my downside of working with AI where it not only gives me poor work work output, but also tells me with very high confidence. [00:26:21] S1: If I would say to you areas that frustrate you, that you feel are inefficient but you don't necessarily know how and where to apply artificial intelligence? Like, what are the things that springs to mind?
00:26:21 · Chaz Englander, David Weisburd
[00:26:21] S1: If I would say to you areas that frustrate you, that you feel are inefficient but you don't necessarily know how and where to apply artificial intelligence? Like, what are the things that springs to mind? [00:26:33] S0: I think video production, editing the first part of our... The podcast. So we have an editor that does a bunch of fancy stuff, but just like the pure editing, the platforms have... Haven't been able to do that. In terms of thinking strategically, like mapping guests, mapping outreach, all of that, that would be a a a big help. Also, obviously, on the asset management side, figuring out which companies we would want to, like, what should be on our target list, mapping the network of how to get to those people and how to get to those companies. These are the things that really take a lot of man hours. [00:27:05] S1: Okay. Excited now. So that's great. And then what I always ask as well is to what extent across those, I'm particularly interested in the kind of more prospecting one as you think about, the managers. What tools of a tool have you used to kind of think about or look at that overall use case?
00:27:05 · Chaz Englander, David Weisburd
[00:27:05] S1: Okay. Excited now. So that's great. And then what I always ask as well is to what extent across those, I'm particularly interested in the kind of more prospecting one as you think about, the managers. What tools of a tool have you used to kind of think about or look at that overall use case? [00:27:21] S0: We're old school. We talk to a lot of other GPs, figure out which companies they like. We don't have a great process for prospecting. We're very much reactive in terms of the conversations that we have. I'd love to be much more proactive. So the answer is we don't really have a process for that outside of just gathering information old fashioned way, meetings, Zoom meetings, in person meetings, and building relationships with co investors. [00:27:43] S1: And then my last question is how do you get the feedback loop of, the areas or the nuggets of things that your listeners are interested in other than people You're asking for [00:27:52] S0: my secret sauce. So I'll I'll give you I'll give you some secret sauce. So we get analytics from the audio, but the richest analytics we actually get from YouTube. And what I started to do about two months ago is we have the retention curves. I take the retention curve, I upload that to AI, I take the transcript, I upload that to AI, and I have it give me feedback in terms of like where did people drop off. So we kinda do this recursive improvement.
00:28:17 · Chaz Englander
[00:28:17] S1: Yeah. Nice. Nice. Well, and then what we try and do if I was working with you as a customer is like, we try and at least kind of whiteboard out, it wouldn't be as quick as this, we would whiteboard out a couple of use cases and try and get feedback on them in real time. Mainly because I think that's what kind of gets both the creative juices flowing of what's possible but directionally where we think AI is good at. So the things that were springing to mind as you speaking by the way are things like, okay well in the comments on YouTube, how are we thinking about categorizing those comments, right? And how are we then thinking about the different nuggets of content off the back of that? So by way of example, have you thought about if I said to you, you could have a workflow that the second that a video is published, it's monitoring those comments. So it's actually a workflow that's connected to that video, and it's in real time grabbing those comments and it's categorizing them, with a view that it's consistently updating this output, sort of summarizing where it feels that thematically, everyone is interested in, or or the main areas of interest.