The Spring 2026 version of this course is no longer satisfied with "explaining how AI impacts the economy"—it now requires students to spend a semester turning an AI startup concept into a working prototype, making venture building itself the pedagogy.
- 01The Stanford Bulletin official course description explicitly states that Spring 2026 will "emphasize venture creation": students must form teams to develop and refine an AI-driven product or startup concept, delivering a "working prototype + venture strategy" at the end of term
- 02The course has been offered for at least three iterations—the Digital Economy Lab website notes that "over the past three years" speakers have included Eric Schmidt, Mira Murati, Jeff Dean, David Autor, Condoleezza Rice, and Reid Hoffman
- 03The only publicly available full syllabus is the Spring 2025 edition: 10 weeks of Tuesday lectures + Thursday discussion sections, with guests including Anthropic co-founder Jack Clark, OpenAI's first chief economist Ronnie Chatterji, MIT's David Autor and Daniela Rus
- 04The 2025 discussion section's VC pricing workshop cited data: in 2024, venture capital invested roughly $130 billion in AI startups
- 05Final grading is divided into five equal components (each 20%): weekly assignments, project progress, final presentation, final documentation, discussion participation—this is the 2025 structure; whether it carries over to 2026 is not yet public
- 06This article is a study guide, not original research—the detailed Spring 2026 guest list has not been made public, and this article does not fabricate any missing information
A cross-department
economics and engineering course
ECON295 / CS323 "The AI Awakening: Implications for the Economy and Society" is cross-listed between Stanford's Economics Department and Computer Science Department, worth 3–4 credits, with Letter grade or Credit/No Credit options, and application-based admission—primarily targeting graduate students in economics, business, computer science, and related fields.
The instructor Erik Brynjolfsson holds a long string of titles: Jerry Yang and Akiko Yamazaki Professor, Senior Fellow at Stanford HAI, Director of the Stanford Digital Economy Lab, Ralph Landau Senior Fellow at SIEPR, Courtesy Professor at Stanford GSB and the Economics Department, and NBER Research Associate. This course is the flagship offering directly run by the Digital Economy Lab (a unit under Stanford HAI) that he directs.
The course features a weekly guest lecture by "leading figures in AI, business, economics, and industry," paired with frontier research discussions and practical applications—this format itself has remained unchanged over three years. What has changed is the Spring 2026 endpoint: from "understanding" to "building."
After three years of
top-tier guests
The Digital Economy Lab course page contains one sentence that is key to understanding this course's history:
"Over the past three years, speakers have included Eric Schmidt, Mira Murati, Jeff Dean, David Autor, Condoleezza Rice, and Reid Hoffman."
This sentence confirms two things: the course has been offered for at least three iterations; and past guests span technology (former Google CEO Eric Schmidt), AI labs (Jeff Dean, former OpenAI CTO Mira Murati), labor economics (MIT's David Autor), policy (former Secretary of State Condoleezza Rice), and industry networks (Reid Hoffman)—a quintessential "top-tier public lecture guest list."
Meanwhile, the Stanford Bulletin official course description for Spring 2026 contains an explicit pivot signal:
"This course examines how advances in AI are transforming the economy and reshaping the frontier of entrepreneurship. In Spring 2026, the course will emphasize venture creation: students will explore how large language models and other AI tools enable small teams to build products and companies with unprecedented speed and scale. […] Working in interdisciplinary teams, students will develop, prototype, and refine an AI-enabled product or startup concept, culminating in a final presentation of a working prototype and venture strategy."
In plain language: for the past three years this course was essentially a lecture series "inviting heavyweights to talk about how AI is changing the economy"; in Spring 2026, the deliverable becomes a real AI startup prototype—interdisciplinary teams develop, refine, and ultimately present a "working prototype + venture strategy" as their final presentation. Spring 2026 applications closed on 2026-03-16 (priority deadline 2026-03-09), indicating this pivot is not a paper plan but an arrangement already implemented at the admissions stage.
Important caveat: the detailed weekly schedule and guest list for Spring 2026 are not currently public. The next section §02 presents the Spring 2025 full syllabus—the only verifiable, actually-run teaching structure available, used as an evidentiary baseline for "how this course was taught in the past," not as a prediction of "how it will be taught in 2026."
The only public
full syllabus
Spring 2025 Econ295/CS323, taught by Erik Brynjolfsson, was structured as 10 Tuesday lectures + Thursday discussion sections + final presentations; the full syllabus PDF is publicly available. This is the economics-lecture format that preceded the "venture creation" pivot—understanding it is essential to seeing what the 2026 pivot actually changes.
Final presentations on June 6: student teams (max 3 per team) present AI startup / policy / research proposals. Grading is divided into five equal components at 20% each: weekly assignments, project progress, final presentation, final documentation, discussion participation—this "team proposal" final format actually foreshadowed the 2026 "venture creation" pivot, except that the 2025 output was an investment-paper-style proposal presentation, while the 2026 official description requires a "working prototype," a higher bar much closer to real startup building.
Three signals from the discussion sections
The Thursday discussion section topics themselves are worth noting: "A Career in AI Research" (research career paths); "How do you pitch an AI startup to VCs?"—a VC pricing workshop, which cited the data point that in 2024, venture capital invested roughly $130 billion in AI startups; "Economic Impact of AI on Labor"—a workshop on macro growth models for automation scenarios. There was also a special discussion guest: Larry Summers. The VC pricing workshop already existed in 2025, showing that "teaching students how to sell AI concepts to investors" didn't materialize out of thin air in 2026—it simply escalated from one discussion module to the course's core deliverable.
If a Chinese university wanted to offer a comparable course, the "invite heavyweights to lecture" part wouldn't be hard to replicate—the hard part is the ecosystem behind the "VC pricing workshop": who prices students' AI startup prototypes, and who provides real investor feedback? The activity level and pace of AI startup financing in China are not on the same order of magnitude as the U.S. AI venture market at the $130 billion level in 2024. Can this course's grading system be transplanted as-is, or does it require a different evaluation framework?
Core readings
spanning three academic lineages
The core readings listed in the Spring 2025 syllabus, full list in the public PDF, span three lineages: AI capabilities per se, labor economics, and AI safety and interpretability.
- ▸Rich Sutton — The Bitter Lesson
- ▸Brynjolfsson & Unger — The macroeconomics of artificial intelligence (IMF)
- ▸Brynjolfsson & Mitchell — What can machine learning do? Workforce implications (Science, 2017)
- ▸David Autor — Why Are There Still So Many Jobs? (2015)
- ▸David Autor — AI Could Actually Help Rebuild the Middle Class (Noema, 2024)
- ▸Narayanan & Kapoor — AI as Normal Technology (2025)
- ▸Acemoglu & Restrepo — Artificial Intelligence, Automation, and Work
- ▸Anthropic — Constitutional AI
- ▸Anthropic — Tracing the Thoughts of a Large Language Model
- ▸Mullainathan & Spiess — Machine Learning: An Applied Econometric Approach
This list itself serves as a useful map: Sutton's Bitter Lesson represents the techno-optimist camp that "believes in compute scale"; Autor / Acemoglu-Restrepo represent labor economics' cautious assessments of automation; Anthropic's two papers wedge safety and interpretability into the required reading of an economics course—forming a stark contrast with MS&E 435 (Series № 01), which has an almost entirely non-overlapping reading system: one reads VC memos, the other reads academic papers and lab safety reports.
Understanding from the outside,
and building from the inside
Series № 01, "MS&E 435: The Economics of the AI Supercycle," and this course, CS323, are two "AI and the economy" courses at the same university, in the same spring semester, with completely different postures.
Placing the two courses side by side reveals a larger signal: two "AI and the economy" courses offered by Stanford in the same semester are both shifting toward "hands-on"—MS&E 435 has students use a computational framework to decompose the industry, while CS323 from 2026 directly requires students to turn frameworks into something that runs. The economics department is no longer satisfied with teaching students "how to analyze the AI economy"; it is now teaching students "how to enter the AI economy and play." This itself is an educational paradigm shift worth recording, not merely a course selection reference.
What we don't yet know
is more important
The hardest discipline in writing this study guide was resisting the urge to guess what the Spring 2026 classroom will actually look like—no one has publicly released the 2026 weekly schedule, guest list, or grading rubric. What can be confirmed is only two layers of evidence: the pivot phrasing "emphasize venture creation" in the Bulletin official course description, and the Spring 2025 full syllabus as an evidentiary baseline for "how this course actually ran in the past."
For readers tracking the Chinese AI education ecosystem, the most worthwhile thing to follow about this course is not the guest list itself, but the shift in the evaluation system: when "explaining AI economics" is no longer the endpoint, and "building a working AI startup prototype" becomes the endpoint, who evaluates this prototype? How does investor feedback get integrated into the classroom? Once this mechanism is running, it may be more worth replicating than the guest list itself.
— Frontier AI in University series · From "understanding the AI industry" to "building an AI company"If you wanted to replicate the 2026 version of this course at a Chinese university, the hardest part wouldn't be securing comparable guests—it would be building the closed loop of "student prototype → real investor feedback → iteration." This mechanism itself may be the course's true pedagogical innovation, not yet another list of luminaries.
This series has 2 articles
Two new Stanford courses in Spring 2026, both covering "AI and the economy," with completely opposite postures: one teaches you to stand outside the industry and use frameworks to understand where the money goes; the other makes you step into the arena and build the company yourself.