AI at 70 Hackathon
Register
1956 2026 Seventy years of artificial intelligence
AI at 70

What must
remain human?

Part of the AI at 70 conference, the hackathon gives students 24 hours to build a working project that takes on the conference's central questions. Organized by the Dartmouth AI Club, with support from Anthropic.

01 — The challenge

Build toward a future worth having.

The most transformative innovations start from big ideas about what's possible. Over 24 hours you'll build something that shows how AI could genuinely contribute to people flourishing — and wrestle out loud with the hard questions that come with it.

A future in which AI helps eliminate disease, expand opportunity, strengthen democracy, and help people find meaning.

The conference asks two questions: Where does AI genuinely improve human decision-making? And where must judgment remain human? Each track is a lens on them. Show your answer in what you build — we want to see your thinking as much as your code.

Every team answers
Who are you building this for, and why do they need it?
Every team answers
What could go wrong, and what would you do about it?
Every team answers
How does this help people rather than decide for them?
02 — Tracks

Five lenses. Pick one.

Tracks are tailored to campus problems. Open one for example projects and the ethical questions judges will press on.

01 Biology & Physical Health Make healthcare more accessible, or help people manage physical health +

Healthcare is expensive, inaccessible, and confusing. Knowledge that could help people is locked behind jargon and paywalls — and even those with access struggle to navigate the system.

Example projects
  • Diagnostic aids for underserved clinics
  • Patient education platforms
  • Medication adherence tools
  • Symptom assessment and triage
  • Health literacy for treatment decisions
  • Accessibility tools for navigating care
Ethics judges will press on
  • Medical accuracy — when does this hand off to a professional?
  • Privacy of deeply sensitive health data
  • Bias in medical AI across populations
  • Transparency about limitations
02 Neuroscience & Mental Health Expand access to mental health support and understanding +

There aren't enough therapists, stigma prevents people from seeking help, and quality care is gatekept by insurance and geography. Many suffer without understanding what's happening to them.

Example projects
  • Evidence-based therapeutic support tools
  • Mental health literacy and psychoeducation
  • Knowing when to seek professional help
  • Stress and burnout prevention
  • Peer support facilitation
  • Tools for neurodivergent students
Ethics judges will press on
  • Crisis situations — when is human intervention required?
  • Privacy and stigma around mental health data
  • Cultural competence; avoiding bad advice
  • Not replacing human connection when it's needed
03 Economic Empowerment & Education Remove barriers to learning and opportunity +

Talent is universal; access to support isn't. Quality education and economic advancement stay gatekept by geography, wealth, and language.

Example projects
  • Adaptive tutoring
  • Career transition guidance
  • Financial literacy for underserved communities
  • Interview prep; scholarship matching
  • Small business and nonprofit support
  • Practical language learning
Ethics judges will press on
  • Who gets excluded — devices, literacy, language?
  • The cost of bad advice about careers or school
  • Informed choices vs. choosing for people
  • Cultural assumptions about "success"
04 Governance & Collaboration Help people participate in democracy, or work together better +

Polarization makes productive dialogue hard, civic participation is down because people feel powerless, and communities can't coordinate around shared challenges.

Example projects
  • Community deliberation platforms
  • Policy explanation and legislative transparency
  • Grassroots organizing support
  • Participatory budgeting
  • Conflict mediation
  • Nonpartisan voter education
Ethics judges will press on
  • How could this be weaponized for manipulation?
  • Whose voices get centered vs. drowned out?
  • Facilitating dialogue without "both-sidesing" reality
  • Privacy in civic participation
05 Creative Flourishing Amplify human creativity and help people find meaning +

As AI handles more routine work, questions about human purpose get urgent. People who could find meaning through creative expression often never get the chance.

Example projects
  • Community storytelling platforms
  • Music and art education with real-time feedback
  • Writing development tools
  • Language and craft preservation
  • Intergenerational knowledge exchange
  • Purpose exploration guides
Ethics judges will press on
  • Humans as creators using tools — not the reverse
  • Enhancing vs. short-cutting the creative process
  • Respectful celebration vs. appropriation
  • Creating or destroying value for creative workers
03 — Run of show

Two days, one build.

Times firm up with the venue. Conference programming runs alongside on Thursday and Friday.

Friday, October 30
~5:30 p.m. Doors, check-in, team matching. Come with a team of up to five or leave with one.
Kickoff Tracks, rules, and judging criteria. Anthropic mentors introduced; clock starts. No commits before start time.
+1 hour Build with Claude orientation. Mark Dominguez leads a short live demo of Claude and Claude Code. Optional if you're already running.
Evening Dinner, then heads-down. Mentor office hours circulating through the room.
Saturday, October 31
Morning Breakfast and final office hours. Last call for technical help.
Midday Submissions close on Devpost. Fifteen minutes of grace, then the form locks. Repo, demo video, and your written thinking.
+60 min Judging at your table. Judges split the room, then reconvene on finalists.
Finals Top three demo live, three minutes each. Strictly timed.
Close Winners announced. The hackathon closes right after lunchtime so you can travel home.
04 — Prizes & judging

Scored out of a hundred points.

Judges are chosen by the hackathon team, led by Dartmouth faculty, and Anthropic staff may join them. A working prototype matters — so does being able to explain who it's for, who could be harmed, and what you'd do about it.

Impact potential 25
Technical execution 30
Ethical alignment 25
Presentation 20
Prizes
  • 01Top three teams: a free, time-limited Claude Max subscription for each team member, plus Anthropic hats.
  • 02First place: also a 30-minute virtual conversation after the event with someone at Anthropic whose work relates to your project.
  • 03All participants: stickers. More prizes will be announced.
What we're looking for
  • Real problems for specific people — not hypothetical problems for "everyone"
  • AI that empowers people rather than replacing or diminishing them
  • Clear thought about who benefits, who might be harmed, and what safeguards apply
  • A working prototype that demonstrates the core idea
05 — Mentors & judges

Anthropic, in the room.

Anthropic colleagues are on site to open the hackathon and help teams as they build. A few more may join as travel allows.

Anthropic Mark Dominguez Leads the Build with Claude orientation that opens the hackathon, and helps teams as they build.
Anthropic Jackson Dolphin On site to help teams as they build.
Anthropic Nathan Eagle Driving up from Boston to help teams as they build. Faculty judges will be announced here.
06 — Partners

Who is behind it.

A student hackathon organized by the Dartmouth AI Club, with support from Anthropic, and part of the AI at 70 conference.

With support from Anthropic People on site to help teams as they build, plus prizes for the top teams and stickers for everyone.
Student organizer Dartmouth AI Club Runs the event: venue, teams, volunteers, submissions, and the run of show.
Host AI at Dartmouth · AI at 70 The "What Must Remain Human?" conference. Full details are on the conference site.
Conference supporter The Rockefeller Foundation Supports the AI at 70 conference.
TBA Campus partners Additional student organizations, labs, and centers joining as partners — announced with the venue.
07 — Before you arrive

Get your Claude account ready.

Bring your own Claude account so you can use features like connectors, skills, and plugins. We can't set up credits or accounts for you, so sort yours out before you arrive.

08 — Questions

The short answers.

Do I need to know how to code?+

No. Strong teams mix builders with people who understand the problem. Chat-first prototypes are welcome, and the Build with Claude orientation on Friday will get you started.

Is there an age requirement?+

Yes. Participants must be 18 or older. Anthropic's terms require Claude account holders to be 18+, which also matches Dartmouth's protection of minors policy.

Can students from other schools take part?+

Yes. The hackathon is open to students from Dartmouth and other institutions. Register and we'll confirm details by email.

Do I need a team to register?+

Register as an individual — it's simpler, and we run team matching at check-in. Pre-formed teams are welcome; just list your teammates on the form. Teams can have up to five people.

Can I bring an existing project?+

New builds only. No commits before the Friday start time — bring ideas, sketches, and research instead.

What do I bring?+

Laptop, charger, student ID, and your own Claude account, ready to use. Meals and snacks are covered. Overnight logistics confirmed with the venue announcement.

How do I submit?+

Through the event's Devpost page: repo link, a short demo video, and written answers to the three critical questions. Link goes out with logistics one week before.