We've been on a long, strange trip together as we enter the final stretch of our AI journey. Even within just the span of the past 3 months we've witnessed first-hand the ground shift beneath our feet with rapid and non-stop advancements in AI research, models, and new products/startups. We've shared a front-row seat to witness the AI transformation of everything from work to the economy, to geopolitics to the way society self-organizes and reassesses our sense of self. It's been fun to get to know a bit about each of you and collectively share our anxieties, fears, curiosity, and excitement about how we can best bend the future arc towards human thriving and away from dystopian technology-based nightmares.
Our final week is entirely focused/graded on producing the first draft of your final semester project. Since the class voted to fold Miniproject #4 into the final project, this will account for 10% + 20% = 30% of your grade in this course. Half of this will be based upon the first draft of your poster due by next week on Wednesday, December 4th by the start of class at 7pm. Unfortunately, there can be no extensions since we're at the end of the semester which is why readings/coding for this week are minimal.
The goal is to produce a complete rough draft while the final poster should be backed by a project/code that realizes your project and produces results with analysis as appropriate. I have office hours Thursday 11-1pm and will hold some early next week by appointment in the afternoon. Ping me via [email protected] if you have any questions, concerns, or comments.
- Use the IPHS200 Poster Template, customize color, layout and images as you like.
- For grading, poster should be submitted in ppt form (if working in a different format, please save as a powerpoint and check formatting)
- No MacOS *.key or pdf files can be accepted (sorry!)
- Abstract: outline project
- Background: cite several prior work/related research
- Methodology: explain tool(s) and methodology
- Results: explain your findings/visualizations
- Your poster should include visuals, analysis, a conclusion, and suggest avenues for future research. There should also be a list of at least 3-5 references.
- Project should not plagiarize by directly copying others’ work or analysis.
- You must search for similar projects (e.g. on kaggle, github, google, etc) and directly reference and cite these.
- Project should be spellchecked and checked for grammatical errors. If you’re unsure, copy and paste text boxes into a word or google document to review.
The FINAL VERSION of your project poster is DUE at the end of the exam period for your class. We are not allowed (per college rules) to accept final projects after the end of the semester (i.e. Friday at 4:30). Again, no wiggle room here.
Here are clear guidance on using external code and resources in your projects.
-
If you build upon existing demos, tutorials, or code, please cite these sources in your references section. This aligns with Kenyon College's academic integrity policies and professional best practices.
-
Your project's value comes from your unique contribution - whether that's an innovative research question, novel datasets, creative prompts, or insightful analysis. Focus on clearly articulating what makes your approach distinctive.
-
If you choose to use no-code or low-code solutions (like the new NotebookLM API), your project will be evaluated on how creatively and effectively you leverage these tools to address meaningful research questions. You can find helpful tutorials for NotebookLM in the Week 13 materials.
As someone who has worked extensively with AI agents since 2019, including projects like the NEH-funded DivaBot collaboration with Denison, I'm here to help you navigate these guidelines and make informed choices about your project approach.
Please don't hesitate to email me if you need any clarification about these guidelines. Your success in this project is important to me.
- OpenAI Sora
- NotebookLM: Low-Code NotebookLM with API: New Features and Customizability to programmatically generate research, scripts or two person conversational podcasts
- Don't Pay $200 Yet! (and more AI Models from Last week)!!! (15:56) (9 Dec 2024)
- (view)(Model Context Protocol (MCP)) Anthropic's New Agent Protocol! (15:34) (29 Nov 2024)
- (PydanticAI)We've Been Building AI Agents WRONG Until Now (19:51) (8 Dec 2024)
- Bee Agent Framework
- a16z ELIZA modelcontextprotocol/servers)
- Multi-Agent AI EXPLAINED: How Magentic-One Works (16:38) (Nov 2024)
- Microsoft JARVIS
- JarviX: A LLM No code Platform for Tabular Data Analysis and Optimization by Liu et al. (3 Dec 2023)
- 2 LLMs Talking With Each Other with Ollama Locally - Xplore TerminaLLM (9:00) (11 Dec 2024)
- Generative Agent Simulations of 1,000 People by Park et al. (15 Nov 2024)
- LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods by Li et al. (10 Dec 2024)
- BREAKING NEWS: oTToDev is Now the OFFICIAL Open Source Bolt.new (11:43) (11 Dec 2024)
- Tony Seba: Greatest disruption in history is here, millions of jobs to disappear (40:30) (Aug 2024)
- David Sacks appointed AI Czar
- Whitehouse FACT SHEET: Biden-Harris Administration Outlines Coordinated Approach to Harness Power of AI for U.S. National Security
- The global AI race: Will US innovation lead or lag? (6 Dec 2024)