Key Moments
Using AI in Research | Stanford CS193T Guest Panel with Mark Algee-Hewitt, Megan Ma & Markus Covert
Stanford OnlineKey Moments
AI is not just for STEM; it's transforming humanities and law. However, its use raises ethical questions about authenticity, authorship, and the potential for misuse, demanding a focus on transparency and human oversight.
Key Insights
AI's ability to analyze vast text datasets allows for 'reading at scale,' identifying patterns imperceptible to human readers, as demonstrated in literary analysis.
The legal field faces challenges in defining quality and value, with AI's impact on billable hours and the potential to automate tasks previously reliant on individual lawyer intuition.
Simulations using AI can create realistic training scenarios for legal professionals, such as witness depositions, allowing practice in high-stakes environments without real-world consequences.
AI's role in bioengineering is advancing complex disease modeling, with the creation of full-cell simulations that predict phenotypic outcomes based on genetic content.
A significant concern in education is that students using AI tools may experience negative psychological impacts like depression and anxiety, alongside poorer learning outcomes.
Transparency in AI use is crucial, especially when claims of authorship or originality are involved, such as in recommendation letters or academic papers.
AI's expanding reach beyond STEM into humanities and law
The panel highlights how Artificial Intelligence is no longer confined to STEM fields but is actively being integrated into humanities and law. In English, AI aids in 'reading at scale,' analyzing vast textual data to identify patterns imperceptible to human readers, and exploring the relationship between style and meaning. The legal field, through initiatives like the 'Lift Lab,' is investigating AI's application in legal practice, focusing on aspects like case assessment, service delivery, and legal education. Bioengineering is leveraging AI for complex biological modeling, such as simulating entire cells to predict outcomes based on genetic data. This cross-disciplinary adoption underscores AI's versatility and its potential to revolutionize research methodologies and outcomes across academia.
Rethinking legal practice and education with AI
Megan Ma from Stanford's Law School discusses how AI is challenging traditional notions of quality and value in the legal profession. The billable-hour model is under pressure as AI can significantly improve productivity, raising questions about revenue models and the exposure of AI in legal work. The lab is exploring AI's role in legal education, developing AI-based simulation training for tasks like witness depositions and negotiations. These simulations aim to bridge the gap between law school learning and practical application, allowing for repeated practice in high-stakes scenarios. The development of AI 'legal partners' or 'digital twins' for senior partners is also being investigated, aiming to capture individual legal styles and decision-making nuances, though current AI tends to lean towards stereotypes.
AI in bioengineering: simulating life and complex systems
Markus Covert from Bioengineering emphasizes the complexity of diseases like cancer, which arise from interactions between numerous genetic, epigenetic, and environmental factors, rather than a single gene. His lab is building full-cell computational models that incorporate genetic content to predict phenotypic outcomes, simulating processes like bacterial growth, DNA replication, and cell division. These simulations generate vast amounts of data and can adapt to environmental changes, leading to altered cell behavior. Covert frames this work as 'artificial life,' distinct from traditional AI, and highlights its intersection with AI, particularly deep learning for image analysis and cell tracking. This advancement is crucial for understanding and treating complex biological systems.
The double-edged sword of AI in education
The panel acknowledges AI's potential to revolutionize education, but also its significant risks. A report from MIT revealed that students using AI tools experienced worse learning outcomes, increased depression, isolation, and anxiety. This has led some educators, like Mark Algee-Hewitt, to refrain from using AI for tasks like writing recommendation letters, prioritizing genuine human connection and quality over mere time-saving. The discussion points to the danger of students outsourcing cognitive abilities, potentially leaving university without acquiring essential skills. The need to balance AI's utility with the preservation of critical thinking and human judgment is paramount, ensuring AI augments, rather than replaces, genuine learning.
Navigating ethical dilemmas and the demand for transparency
A core concern throughout the discussion is the ethical use of AI and the need for transparency. When AI is used for tasks like writing recommendation letters or academic papers, the lack of disclosure can lead to a loss of personal voice and trust. The legal profession sees young lawyers hesitant to use AI due to high-profile cases of AI generating false legal citations, creating tension with partners who encourage its use. Mark Algee-Hewitt proposes radical transparency as the solution, advocating for clear disclosure of AI usage in academic and professional contexts to foster dialogue and allow for informed opinions. This transparency is seen as essential for developing societal norms and trust around AI.
Defining essential human skills in the age of AI
Panelists identified critical human skills that remain indispensable despite AI advancements. Critical thinking was highlighted as paramount, enabling individuals to interpret information, rely on their own intellect and experiences, and avoid over-reliance on AI for cognitive tasks. In law, 'judgment' was emphasized, involving the ability to build a necessary reference library of experiences and to craft narratives. The concept of 'scientific taste'—knowing what problems are worth solving and how to make a unique impact—was also mentioned. The ability to think critically, engage in deep reflection, collaborate with others, and possess intellectual 'dual-bravery' to combine knowledge from different fields are seen as vital for navigating a future where AI handles routine tasks.
AI as a tool for augmentation, not replacement
The conversation consistently circles back to AI's role as a tool for augmentation. Rather than having AI perform tasks entirely, the focus should be on how it can enhance human capabilities. For example, AI can be used to generate creative prompts or provide alternative perspectives, pushing human users to think more deeply and produce higher-quality work, as seen in the example of writing research paper paragraphs. In education, AI can facilitate game design by handling mechanical aspects, allowing students to focus on the creative process. The underlying principle is to leverage AI to extend human potential, encouraging curiosity and deeper engagement, rather than allowing it to diminish our own cognitive abilities.
The future of AI: a 'seismic shift' and the importance of human ingenuity
The panelists agreed that the current AI revolution is just the beginning, potentially a 'seismic shift' compared to earlier technological advancements like the rise of Google. They stressed the importance of embracing AI, learning to use it effectively, but crucially, not at the expense of developing fundamental human skills. The goal is to use AI to tackle previously insurmountable problems, fostering radical new ideas and communication. The ultimate aim is to ensure AI empowers humanity to solve complex challenges, inspire innovation, and create new connections, rather than simply providing shortcuts or replacements for human intellect and effort. This requires a continued focus on human curiosity, critical thinking, and a commitment to pushing the boundaries of what's possible.
Mentioned in This Episode
●Products
●Software & Apps
●Companies
●Organizations
●Books
●Concepts
●People Referenced
Common Questions
AI is employed in humanities research for tasks like analyzing large volumes of text to identify patterns, conducting in-depth analysis of specific texts, and studying patterns in correspondence networks. It also serves as a subject of analysis, particularly concerning AI-generated writing styles and their impact on meaning.
Topics
Mentioned in this video
Professor of English and Digital Humanities, discussing AI in humanities research, particularly style and meaning in academic writing.
Executive Director of Lift Lab at the Law School, focusing on applied AI in legal practice, including evaluation, service delivery, education, and advanced methods.
Professor and Chair of Bioengineering, discussing AI in understanding complex biological systems, cell simulation, and artificial life.
Known as 'Hash', worked on the first Tensor Processing Units (TPUs) at NVIDIA, crucial for AI advancements like AlphaGo.
Nobel laureate whose work in super-resolution microscopy was mentioned as an example of solving previously intractable scientific problems.
Academic Director involved in formulating AI policies for a research lab, emphasizing understanding the design choices behind AI tool usage.
A dimensionality reduction technique used in Mark Algee-Hewitt's research to visualize topic clusters in academic articles.
An AI model mentioned in the context of its capabilities in writing and as a conversational partner.
A legal AI provider mentioned in the context of how law firms are using AI tools.
An AI tool mentioned in the context of generating code and potentially being used in research settings.
A legal AI provider mentioned in the context of how law firms are using AI tools.
An AI program that defeated Go champion Lee Sedol, enabled by advancements in TPUs.
Gene-editing technology mentioned by Markus Covert as a significant advancement comparable to AI's transformative potential.
An early search engine that, along with others like Lycos and Ask Jeeves, was overtaken by Google.
An early search engine that, along with others like Lycos and WebCrawler, was overtaken by Google.
A platform developed by Michael Bernstein's lab for creating digital twins of individuals, discussed in the context of legal 'partner personas'.
A research and development facility where Markus Covert consulted on artificial life projects.
A company whose leaders met with a Vatican representative to discuss AI and its implications for humanity.
Company where Jonathan Ross works and developed early TPUs, essential for AI computation.
A law firm whose attorneys were involved in a case where fabricated quotes were included in legal filings, highlighting AI-related risks.
An early search engine that, along with others like WebCrawler and Ask Jeeves, was overtaken by Google.
A research center run by Mark Algee-Hewitt that employs students for projects in data science for textual and cultural analysis.
Massachusetts Institute of Technology, cited for a report on AI and education that linked AI tool usage to negative student outcomes.
Legal Innovation through Leading Technology Lab at the Law School, focusing on applied AI in legal practice.
A newspaper that published an article about AI, post-humanism, and the Vatican's perspective.
A test of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human. Discussed as a foundational concept in AI and its ethical implications.
Google's algorithm that revolutionized search engine results, leading to the decline of earlier search engines.
More from Stanford Online
View all 175 summaries
77 minRAG Explained: From BM25 to Agentic RAG | Stanford CS329Z Lecture 3 with Diyi Yang
64 minStanford CS193T Thinking with AI | Autumn 2026 | Managing Sources
77 minStanford CS193T Thinking with AI | Autumn 2026 | Core Prompting
78 minLLMs for Agent Builders: Context Engineering | Stanford CS329Z Lecture 2 with Michael Ryan
Ask anything from this episode.
Save it, chat with it, and connect it to Claude or ChatGPT. Get cited answers from the actual content — and build your own knowledge base of every podcast and video you care about.
Get Started Free