AI is changing academic research and surpassing PhD scholars–Here’s how

The age of artificial intelligence is upon humanity. No domain ranging from finance, and the labour market to security & strategy and banking, is not spared of disruptions caused by AI.

Now, the future of academic research is at stake due to generative AI. Recent concerns over the potential collapse of research were voiced by Alexander Kustov, who is an associate professor studying public opinion on immigration at the University of Notre Dame.

Kustov published a Substack piece outlining the different ways AI tools are upending academia from drafting publishable paper in first-quartile journals to performing peer reviews with more efficiency.

Taking to X, Kustov wrote, “AI can already do social science research better than most professors with PhDs. And, for the first time in my life, I really have no idea what will happen in five years.”

“Things are changing already, we just need to wake up,” he added.

Here are the different ways in which AI models are outperforming most social science PHDs.

AI is beating professors at research

The artificial intelligence models including Anthropic Claude are capable of surpassing the social science professors at research. It is a mistake to assume that such capability will mark the end of “human-only superiority” in academia.

According to Kustove, AI can write high-quality research papers that meet the standards of top-tier journals. Consequently, the cost of manuscript creation has plummeted to roughly $100 and required only a few hours of prompt engineering. A few examples validate this argument.

Tibor Rutar, assistant professor of sociology at the Faculty of Arts (University of Maribor) wrote a publishable-quality research paper by using AI prompts alone.

Similarly, Yascha Mounk, a German-American political scientist and author, also claimed that Claude can produce high-quality political theory papers in just a few hours.

Death of “Artisanal” research paper

The traditional 30-page academic paper is now becoming obsolete in the age of AI, turning it into “vestigial wrapping paper.”

AI now matches the capabilities of professors in synthesizing ideas, conducting literature reviews and generating top-quality manuscripts.

Sean J. Westwood, associate professor in the department of Government at Dartmouth College, said, “AI does lit reviews better. AI will do peer review. Users will skim AI summaries. The real science is the question, the pre-analysis plan, and the analysis.

Hence, AI has changed the role of researcher from executor to one who focuses solely on providing accurate “inputs” to models.

Imminent collapse of commercial journal system

AI models are making the economic model of academic publishing unsustainable.

If AI slashes the cost of generating a paper, the submission rates will spike, leading to a 90 percent desk-rejection rate. The model will also create a review crisis, replacing human-led peer review systems and monopolizing the AI-assisted screening and post-publication review models.

Paradigm shift in roles and skills

Owing to the rise of agentic AI, the position of traditional research assistant and “apprenticeship” model are under threat. For instance, AI agents can perform technical grunt work, such as data cleaning and regression running.

For junior scholars, the new premium is on “original thinking” and “verification.”

Silver lining

According to Kustov, the risks posed by AI to academic research are real, but these models also offer democratization of science.

AI enables highly interactive and adaptive surveys and offers real-time translation of research into policy solutions. It also promotes global equity in science by breaking the monopoly of elite Western universities. Now non-native English speakers can generate high-quality prose that reads as well as anything produced by Cambridge or Stanford.