Are Economists Open to AI? Text as Data as Survey on Professional Sentiment and Academic Research Trends
This paper introduces TaDaS, a novel framework that transforms unstructured internet text into survey-like evidence to analyze economists' professional sentiment toward AI, revealing that while initial reactions are negative, increasing visibility of AI in elite journals correlates with more favorable attitudes across multiple dimensions.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
The Big Idea: Turning "Chaos" into a "Survey"
Imagine you want to know how a group of people feels about a new, scary technology (like AI). The traditional way to find out is to send out a survey. But surveys have problems: they are expensive, people might lie to look good, and you can't ask questions about things that haven't happened yet.
The authors of this paper say, "Wait a minute. People are already talking about this online!" But there's a catch: online chatter is messy. It's full of jokes, rants, off-topic stories, and noise. It's not a clean survey.
So, the authors invented a new tool called TaDaS (Text as Data as Survey). Think of TaDaS as a smart translator. It takes messy, unorganized internet text and translates it into a structured, reliable "survey" without ever asking anyone a single question.
How TaDaS Works: The "Library and the Coffee Shop" Analogy
To understand how they did this, imagine two different places:
- The Library (The "Answer" Corpus): This represents elite academic journals. It's quiet, organized, and contains the "official" research topics that experts agree are important right now. It's like a map of where the profession is going.
- The Coffee Shop (The "Question" Corpus): This represents EJMR (Economics Job Market Rumors), an anonymous online forum where economists chat, complain, and debate. It's loud, chaotic, and full of raw emotion.
The TaDaS Process:
- Step 1: Drawing the Map. The researchers use the "Library" to draw a map of current topics. They identify specific areas on the map, like "AI," "Gender Economics," or "Health Economics."
- Step 2: Listening to the Coffee Shop. They take the messy conversations from the "Coffee Shop" and use a computer brain (AI) to figure out which part of the map each conversation belongs to. If someone is talking about machine learning, the system tags it as "AI."
- Step 3: Scoring the Mood. Once the conversation is tagged, the system reads the tone of the reply. Is the person curious? Are they angry? Are they being mean (toxic) or confused? It gives a score from 0 to 1 for six different feelings: Openness, Negativity, Toxicity, Arrogance, Curiosity, and Confusion.
What They Found: The "Stormy Start, Sunny End" Pattern
The researchers asked a simple question: Are economists open to AI?
They found a two-part story:
- The Static View (The Snapshot): If you look at a random conversation about AI right now, it tends to be grumpier than conversations about other topics. People are more negative, more arrogant, and less curious when AI comes up compared to, say, "Labor Economics." It's like a stormy day in the coffee shop.
- The Dynamic View (The Movie): Here is the surprise. As AI starts appearing more in the "Library" (the elite journals), the mood in the "Coffee Shop" actually gets better.
- When AI is a hot topic in top journals, the forum discussions become more open and more curious.
- At the same time, the discussions become less negative, less toxic, less arrogant, and less confused.
The Analogy: Imagine a new, scary gadget is introduced. At first, people in the coffee shop are yelling, "It's a scam! It will ruin us!" (High negativity). But as the gadget starts appearing in the official manuals and news (the Library), the yelling stops. People start asking, "How does this work?" and "Can I use this for my project?" (High curiosity). The tension lowers as the technology becomes a normal part of the landscape.
Why This Matters
The paper proves that you don't need to pay people to fill out surveys to understand their feelings. By using this "Text as Data as Survey" method, you can look at existing archives of text and see how a profession is reacting to change in real-time.
Key Takeaways:
- Method: They built a system to turn messy internet comments into clean, comparable data.
- Result: Economists initially react to AI with resistance and negativity.
- Trend: As AI becomes more visible and accepted in top academic journals, the resistance fades, and the conversation becomes more constructive and curious.
- Timing: This shift happens before everyone starts using the tools personally; it's driven by the technology becoming "legitimate" in the eyes of the profession's leaders.
In short, the paper shows that when a new technology enters a professional world, it starts with a fight, but as it becomes part of the official "rulebook," the fighting stops, and the learning begins.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.