AI-generated podcast | Investment giants in the AI age
I recently published a paper titled "Asset Manager Capitalism and the Political Economy of Artificial Intelligence" in Review of International Political Economy.
I used NotebookLM to create an AI-generated podcast about the paper and Otter.ai for the transcript (see below). It turned out pretty cool—give it a listen. 🎧🤖
This experiment taught me a lot—not just about this wonderful AI tool, but also (and more importantly) about the evolving role of academics in communicating research to broader audiences.
By using generative AI to script and narrate the podcast, I was able to distill a complex argument about asset management giants and AI-driven market power into a format that is accessible and engaging for listeners beyond academia.
But this process also prompted deeper reflection: What do we lose when those seemingly thrilled AI-generated voices are presenting our human knowledge? Moreover, what will happen if and when AI does not even bother to ask us whether we want podcasts about our papers? Who will control the voice of scholarly dissemination in such situations?
I was inspired by a 2023 paper published by Ian McCarthy and Marcel Bogers in Business Horizons and titled “The open academic: Why and how business academics should use social media to be more ‘open’ and impactful.” This paper is a call for business scholars to embrace social media to bridge the research-practice gap. The rise of generative AI tools like NotebookLM made me wonder how to maximize the benefits of AI-generated media as a form of research dissemination through social media platforms while avoiding bad practices.
McCarthy and Bogers (2023: 160) present a useful framework about the various approaches to social-media-enabled academic openness, including some “do’s” and “don’ts.” In my case, I am now being a “promoter” because I am promoting my own research to a wider non-academic audience (at least I am trying). According to McCarthy and Bogers, one of the dont’s for academics promoting their research is to substitute “human engagement with robotic, automated interactions that can be spam-like and pushy” (McCarthy and Bogers, 2023: 160). So, I am wondering, is my AI-generated podcast spammy and pushy? This is particularly important because I have not curated a single sentence of that AI-generated podcast.
I see AI-generated podcasts becoming part of a broader movement toward making academic research more open and impactful. It is a useful extension to be explored through the lens of McCarthy and Bogers’ framework. However, I believe that humans should curate the product—something which I have not done at all in my AI-generated podcast as it is just a test—ooops sorry!
To use Albert Borgmann’s (1984) categories, I believe that human curation is necessary to escape the clutches of the “device paradigm”—the tendency of modern technology to make things easier while simultaneously disconnecting us from “focal things”: those meaningful practices (like creating our own podcast instead of relying on AI) that foster a richer human experience.
Would love to hear from others experimenting with AI-generated podcasts for academic outreach. How are you navigating these new affordances?
--
Transcript (generated with Otter.ai):
Welcome back to the deep dive. This time we're gonna we're gonna dive into the world of asset management and AI. So imagine a world where the decisions that are impacting your investments, your financial future are increasingly made by these these algorithms. That's the world we're exploring today. Yeah, it's a fascinating and rapidly evolving landscape, for sure, it really is. And our guide for this deep dive is an academic article. It's called asset manager capitalism and the political economy of artificial intelligence, and it's by Andrea Lagna, and it was published in the review of international political economy journal just last month. And Lagna takes a takes a really interesting approach, I think, instead of simply looking at AI as just like a cool new tool for finance. He really dives into these power dynamics, exploring how AI is reshaping that competitive landscape and potentially shifting power towards a select few. Oh, okay, so let's unpack that a bit. Lagna identifies three key areas to understand this shift. First, how different asset managers, from hedge funds to robo advisors are using AI based on their business models. Second, the specific AI technologies and data they're relying on. And third, the power dynamics emerging between these firms and big tech in the data economy. Yeah, I think each these areas offers a unique perspective on how AI is changing the financial world, and ultimately, you know how your money is being managed. Okay, so let's start with the first area, AI, and the different ways asset managers are doing business. Now, I know that the asset management sector is incredibly diverse. You've got your hedge funds making high risk bets, your mutual funds offering a more stable approach, and then you have robo advisors that are trying to make investing accessible to to everyone. How is AI shaking things up across these different models. Well, you know, imagine a hedge fund like like Renaissance technology, known for their incredibly secretive and successful trading strategies. They've been using AI for years. And, you know, one way they leverage its power is by analyzing massive data sets to identify these, these subtle patterns that humans might miss. So it's not just about replacing human traders with robots. It's about giving them a like a super powered tool to amplify their insights Exactly, exactly. It's about finding those hidden correlations, those tiny signals in the in the noise, that can give them an edge in the market. For example, some hedge funds use AI to analyze social media sentiment, news articles and even satellite images, and they're trying to gage consumer behavior and predict stock movements. Wow, that is. That's mind blowing. So while hedge funds are using AI to gain a competitive advantage, what about robo advisors? They seem to be focused on a completely different goal, right? Yeah. Robo advisors, like Wealthfront are using AI to personalize investment recommendations and really make the process more user friendly, more accessible, they might analyze your risk tolerance, your financial goals and even like your spending habits, and use that to create this, this tailored portfolio. So if I'm using a robo advisor, should I be worried that I'm not getting access to the same sophisticated AI strategies as those edge funds? Or is it a trade off, personalized service for potentially less aggressive returns. It's a great question, and it highlights a crucial point. I think AI in asset management isn't one size fits all. It's about aligning the technology with the specific goals and business models of each firm. So while a hedge fund might be looking for those tiny, hidden signals to make quick profits. A robo advisor is using AI to provide like, a smoother, more personalized experience for individual investors, precisely, and that's where the concept of strategic alignment comes in. Some firms might be overestimating AI's potential or struggling to implement it effectively. Yeah, just because you have the technology doesn't mean you know how to use it to achieve your specific goals. Yeah. It seems like the key is not just having AI but but knowing how to use it strategically. This makes me think about all the technology behind the scenes. So let's shift gears a bit and dive into that second area along the highlights, the data, the infrastructure, that whole black box of AI. Yeah, this is where we get to peek under the hood a little bit see what's really powering these AI systems. I'm ready to to get technical So, so what is fueling all this AI magic in in asset management? Well, you know, at the heart of it all is, is data. Asset managers have always relied on traditional data sources, you know, stock prices, economic indicators, company financials, that kind of thing. But AI is kind of pushing them toward, toward what's called alternative data. Think social media sentiment, you know, satellite images of shipping traffic, even online reviews, anything that might, that might give them an edge in predicting these market movements. So it's like AI is turning the the entire world into one giant data source for these investors. I'm trying to wrap my head around how AI could use something like like social media sentiment to actually make investment decisions. Can you walk me through that? Absolutely. So let's say a company is about to release a new product, AI can analyze millions of tweets, posts and reviews to kind of gage that public opinion and then try to predict whether that product will be.