Course Syllabus

Safeguarding the Future - Spring 2026 Syllabus (MAS.342/842)

 

Feb 4   

Feb 11     

Feb 18     

Feb 25    

Mar 4    

Mar 11  

Mar 18   

Mar 25   

Apr 1   

Apr 8     

Apr 15  

Apr 22

Apr 29

May 6     

What deserves protection / The present cannot endure (Esvelt, Specter)

The history and future of human agency (Crawford)

AI and human flourishing (Pataranutaporn)

Positive visions of the future (Erkers)

Nuclear game theory (Weber)

Rewriting life (Esvelt)

Project work for positive visions <Harvard spring break>

<MIT spring break>

Project presentations / Communicating technical subjects (Specter)

Aligning transformative AI (tbc)

Civilizational resilience and the tech tree (Esvelt) / Raising the wellness waterline (tbc)

Coordinating to solve problems (Gellin)

Class choice or surprise (influential result of 2023's class) 

Final project presentations

Meeting Time: 10:00am to 12:55pm Room: MIT Building E15-341 

Enter using the Ames St at Amherst St door, up the stairs and past the third floor atrium.

Note “MIT time”: class will begin promptly at 10:05am.

 

Class guidelines:

No laptops open during class

Assignments will include two essays in the op-ed style as well as two projects

The mid-semester project: relevant to cultivating and safeguarding a positive vision

The final project: any topic relevant to protecting us from catastrophe (as you define it)

 

Syllabus dates and assignments are subject to change with guest speaker availability

 

February 4

What deserves protection / The present cannot endure      Kevin Esvelt and Michael Specter

At our current rate of economic growth, we’d be using every atom in the galaxy in less than 10,000 years… which is a problem because the Milky Way is more than 50,000 light-years across. In under 150 years, we’ve gone from fountain pens to tweets, cavalry to ICBMs, and snake oil to CRISPR-mediated T-cell therapy … yet somehow, we imagine that the future will consist of people like us, living much the way we do today. If the present world cannot endure without catastrophe or radical transformation, then the future will be quite different from what most people expect. It might offer delight and fulfillment literally beyond our current capacity to imagine, a hellish dystopia, or an empty universe. If we prefer the first option, we’d best figure out how to nudge our trajectory in a positive direction. Are we moving towards a radically different future? What are the crucial drivers? And how can we communicate problems and solutions to best safeguard the future?

 

February 11

The history and future of human agency             Guest: Jason Crawford, Roots of Progress

Humanity has gained increasing control over the essentials of life and the economy: our food supply, our materials and manufacturing processes, our use of energy, and our transportation networks. A prerequisite of safeguarding the future is the continuation of this historical process. Commentators from Samuel Butler to Vernor Vinge to the present have expressed concern that humanity will lose control. We will approach this topic first historically, arguing the apparently unpopular position that human agency has been increasing, not decreasing, over time. We will cover present evidence for this view, and suggest key drivers of the trend. Then we will discuss whether we can expect this trend to continue or whether it might reverse, and why.

Required reading:

Recommended reading:

 

February 18

AI and human flourishing  Guest: Pat Pataranutaporn, MIT

The default narratives about AI and humanity tend toward extremes: utopian transcendence, dystopian subjugation, or extinction. But before any of that can happen, if it happens, we have to live through Wednesday. What does it look like for AI to make Wednesdays better? Worse? Meaningful or hollow? How might AI companions and collaborators actually affect human wellbeing, creativity, and connection - not in theory, but in practice? What does AI do to us now? Can human flourishing be promoted by increasing our dependence on systems we don't fully understand? Or is that a natural continuation of civilization? Is there a version of AI assistance that enhances agency rather than eroding it? 

Required reading:

Recommended listening:

 Due: Writing #1

 

February 25

Positive visions of the future Guest: Beatrice Erkers, Foresight Institute

It's easier to imagine the end of the world than a future worth wanting. Dystopias write themselves; utopias tend toward the bland or the sinister. Yet if we can't articulate what we're safeguarding - not just what we're safeguarding against - we're navigating without a destination. What would a genuinely good future look like? Not merely the absence of catastrophe, but the presence of something worth having. Why are positive visions so scarce, why do the ones we have so often fail to persuade, and what it might take to imagine futures that are neither naive nor dystopian in disguise? The goal isn't consensus on a single vision, but developing the capacity to think seriously about where we might want to go.

Required reading:

Recommended reading:

  • The Culture novels (I. Banks) (we recommend starting with The Player of Games)
  • Diaspora (G. Egan)

 

March 4

Nuclear game theory                         Guest: Andrew C. Weber, Council on Strategic Risks

Leo Szilard, the inventor of the chain reaction, warned that dropping the bomb on Japan in 1945 would accelerate an arms race that would make American cities vulnerable to foreign militaries for the first time in half a century. He was right. Nuclear weapons cannot be blocked and can only be deterred through mutually assured destruction. Whether they kept the Cold War cold is debatable, but it was always touch and go. We arguably came close to losing half of humanity in 1983. What can we learn from our first experience with a technology capable of threatening so many of us? And what do the war between Russia and Ukraine and rising tensions between the United States and a China that is modernizing its arsenal imply for global nuclear risk?

Required reading: 

  • NUKEMAP (A. Wellerstein) [instructions: try out some parameters]

Recommended reading:

 

 

March 11

Rewriting Life                                                                                                    Kevin Esvelt

COVID-19 was caused by a single, comparatively mild pandemic virus that wasn’t even engineered to cause harm. Around thirty thousand people today could single-handedly assemble pandemic influenza viruses from synthetic DNA using cheap equipment. For years, scientists have tried to find viruses in nature that can harm us. Should they be doing that? What would happen if they were to post a list of credible pandemic viruses found in nature or describe how to engineer more destructive ones? Can we avoid this while still ending disease and aging? How can we best immunize civilization against catastrophic biology… and delay disaster until we have succeeded?

Required reading: 

Recommended reading:

Due: Writing 1 Re-write

 

March 18

Project Day

To safeguard the future, we can't simply learn from the past and discuss the challenges of the present. We need to act. That may involve creating compelling videos or conducting class exercises that are turned into manuscripts, covered in Science, and help catalyze the new field of AI/biosafety research and the Executive Order on AI. Accordingly, today will focus on project planning. Also, everyone from Harvard will probably be out today.

 

March 25

No class; MIT spring break

 

April 1

Project Day + Communicating technical subjects with minimal controversy             Michael Specter

Present and discuss projects reminding us why we’re doing this. 

The technical challenges we will have to adopt - or choose to reject -will make it essential that technologists express themselves plainly, directly and in ways that non-scientific people do not find threatening. If you can’t express a philosophical point of view in a way that is widely understood how can you hope to set an agenda for a technological future?

Required reading:

Due: Mid-term projects    

Due by the end of class: Final project proposal.

                                       

 

 

April 8

Aligning transformative AI                              Guest speaker: tbd

We have already created systems with narrowly superhuman capabilities. At some point, AI will noticeably accelerate the pace of technology development. That will likely return the world to a cycle of exponential growth and innovation that stalled when we stopped having as many babies as we could feed, interrupting the sequential conversion of inventions into calories into inventors. The resumption of that cycle will mark the end of our current world. When will it happen, and how can we begin to prepare? ChatGPT – while not in any way challenging the dominance of humanity – seemed to wake up many people to the fact that AI will have abilities we have not yet imagined or planned for.

What will happen when we create agents more capable than we are? It's concerning that chimpanzees are nearly extinct and mostly reside in zoos. Anything with tremendous power must be much more aligned with our interests than the average human. And even given perfect alignment, machines with significantly superhuman competence will still upend the strategic gameboard, potentially creating incentives for people to act dangerously or even catastrophically. How might we align transformative AI systems and prevent their capabilities from causing harm? 

Required reading: 

Recommended reading:

Due: Writing #2

 

April 15

Civilizational resilience and the tech tree (Esvelt) / Raising the wellness waterline (tbc)

Are we farsighted, wise, and coordinated enough to navigate increasing technological power? How might we improve our odds? One option: evaluate whether a new advance favors stability or disruption long before it’s made… and pursue defenses immediately. 

Meanwhile, we must live our best, because every day matters. Yet at any given time, an estimated 20% of people experience some symptoms of mental illness. Suffering and mental illness make it hard to cooperate, so helping people lets them help more. What are the most effective treatments today? What can we invent to scalably improve well-being?

Required reading: 

 

April 22

Coordinating to solve problems                                                                        Guest: Bruce Gellin

Our world is governed by a constellation of bodies: governments, industry, academia, to name a few. They need to work together and they rarely do. The COVID pandemic was the most recent and horrifying example of government bodies competing with each other (even just within the United States). Between the ability of government, the desires of industry, and the often theoretical goals of academic researchers, we need to shape how we develop, adopt, and govern use of technologies. How do we corral these forces together? What are historic examples of large coordination efforts, and what lessons from them can we apply today?

Required reading:

 

April 29

TBD depending on class interest and tech advances in 2026   Kevin Esvelt and Michael Specter

Required reading:

 

May 6

Final Project Presentations

 Due: Final projects