The movies embedded on this publish are finest considered on steveblank.com
Worldwide Coverage college students will probably be spending their careers in an AI-enabled world. We wished our college students to be ready for it. For this reason we’ve adopted and built-in AI in our Stanford nationwide safety coverage class – Know-how, Innovation and Nice Energy Competitors.
Right here’s what we did, how the scholars used it, and what they (and we) discovered.
Know-how, Innovation and Nice Energy Competitors is a global coverage class at Stanford (taught by me, Eric Volmar and Joe Felter.) The course gives future coverage and engineering leaders with an appreciation of the geopolitics of the U.S. strategic competitors with nice energy rivals and the function crucial applied sciences are taking part in in figuring out the end result.
This course contains all that you’d anticipate from a Stanford graduate-level class within the Masters in Worldwide Coverage – complete readings, visitor lectures from present and former senior coverage officers/specialists, and deliverables within the type of written coverage papers. What makes the category distinctive is that that is an experiential coverage class. College students type small groups and embark on a quarter-long challenge that acquired them out of the classroom to:
- choose a precedence nationwide safety problem, after which …
- validate the issue and suggest an in depth resolution examined in opposition to precise stakeholders within the know-how and nationwide safety ecosystem
The category combines a number of educating instruments.
- Actual world – College students labored in groups on actual issues from authorities sponsors
- Experiential – They get out of the constructing to interview 50+ stakeholders
- Views – They get coverage context and insights from lectures by specialists
- And this 12 months… Utilizing AI to Speed up Studying
Rationale for AI
Utilizing this quarter to introduce AI we had three issues going for us: 1) By fall 2024 AI instruments have been good and getting exponentially higher, 2) Stanford had arrange an AI Playground enabling college students to make use of a wide range of AI Instruments (ChatGPT, Claude, Perplexity, NotebookLM, Otter.ai, Mermaid, Lovely.ai, and many others.) and three) many college students have been utilizing AI in courses nevertheless it was often ambiguous about what they have been allowed to do.
Coverage college students must learn reams of paperwork weekly. Our hypotheses was that our pupil groups might use AI to ingest and summarize content material, determine key themes and ideas throughout the content material, present an in-depth evaluation of crucial content material sections, after which synthesize and construction their key insights and apply their key insights to unravel their particular coverage drawback. They did all that, and far, a lot, extra.
Whereas Joe Felter and I had arm-waved “we have to add AI to the category” Eric Volmar was the true AI hero on the educating group. As an AI energy person Eric was most frequently forward of our college students on AI expertise. He threw down a problem to the scholars to repeatedly use AI creatively and advised them that they might be graded on it. He pushed them laborious on AI use in workplace hours all through the quarter. The outcomes beneath communicate for themselves.
In case you’re not acquainted with these AI instruments in follow it’s price watching these one minute movies.
Crew OSC
Crew OSC was attempting to know what’s the applicable degree of economic threat for the U.S. Division of Protection to offer loans or mortgage ensures in know-how industries?
The group began utilizing AI to do what we had anticipated, summarizing the stack of weekly coverage documentsusing Claude 3.5. And like all groups, the surprising use of AI was to create new leads for his or her stakeholder interviews. They discovered that they may ask AI to present them an inventory of leaders that have been concerned in related packages, or that have been concerned of their program’s preliminary levels of improvement.
See how Crew OSC summarized coverage papers right here:
In case you can’t see the video click on right here
Claude was additionally in a position to create an inventory of leaders with the Division of Power Title17 credit score packages, Exim DFC, and different federal credit score packages that the group ought to interview. As well as, it created an inventory of leaders inside Congressional Funds Workplace and the Workplace of Administration and Funds that may be capable of present insights. See the demo right here:
In case you can’t see the video click on right here
The group additionally used AI to transcribe podcasts. They seen that key leaders of the organizations their drawback got here from had produced podcasts and YouTube movies. They used Otter.ai to transcribe these. That offered further context for after they did interview them and allowed the group to ask insightful new questions.
In case you can’t see the video click on right here
Notice the ability of fusing AI with interviews. The interviews floor the data within the groups lived expertise.
The group got here up with a use case the educating group hadn’t considered – utilizing AI to critique the group’s personal hypotheses. The AI not solely gave them criticism however supported it with hyperlinks from revealed students. See the demo right here:
In case you can’t see the video click on right here
One other use the educating group hadn’t thought was utilizing Mermaid AI to create graphics for his or her weekly displays. See the demo right here:
In case you can’t see the video click on right here
The surprises from this group saved coming. Their final was that the group used Lovely.ai as a way to generate PowerPoint displays. See the demo right here:
In case you can’t see the video click on right here
For all groups, utilizing AI instruments was a studying/discovery course of all its personal. By and huge, college students have been largely unfamiliar with most instruments on day 1.
Crew OSC steered that college students ought to begin utilizing AI instruments early within the quarter and experiment with instruments like ChatGPT, Otter.ai. Instruments that which have steep studying curves, like Mermaid must be began on the very begin of the challenge to coach their fashions.
Crew OSC AI instruments abstract: AI instruments are usually not excellent, so be sure that to cross verify summaries, insights and transcriptions for accuracy and relevancy. Be actually crucial of their outputs. The largest takeaway is that AI works finest when ready with human efforts.
Crew FAAST
The FAAST group was attempting to know how can the U.S. enhance and scale the DoE FASST program within the pressing context of nice energy competitors?
Crew FAAST began utilizing AI to do what we had anticipated, summarizing the stack of weekly coverage paperwork they have been assigned to learn and synthesizing interviews with stakeholders.
One of many options of ChatGPT this group appreciated, and vital for a nationwide safety class, was the short-term chat function – knowledge they entered wouldn’t be used to coach the open AI fashions. See the demo beneath.
In case you can’t see the video click on right here
The group used AI do a couple of new issues we didn’t anticipate – to generate emails to stakeholders and to create interview questions. In the course of the quarter the group used ChatGPT, Claude, Perplexity, and NotebookLM. By the top of the 10-week class they have been utilizing AI to do a couple of extra issues we hadn’t anticipated. Their use of AI expanded to incorporate simulating interviews. They gave ChatGPT particular directions on who they wished it to behave like, and it offered personalised and customized solutions. See the instance right here.
In case you can’t see the video click on right here
Studying-by-doing was a key a part of this experiential course. The large concept is that college students be taught each the strategy and the subject material collectively. By studying it collectively, you be taught each higher.
Lastly, they used AI to map stakeholders, get recommendation on their subsequent coverage transfer, and requested ChatGPT to evaluation their weekly slides (by screenshotting the slides and placing them into ChatGPT and asking for suggestions and recommendation.)
The FAAST group AI software abstract: ChatGPT was particularly good when utilizing photographs or screenshots, so in these multi-level duties, and once you wished to make use of type of extra customized directions, as we used for the stakeholder interviews. Claude was higher at extra conversational and human in writing, so used it when sending emails. Perplexity was higher for researchers as a result of it gives citations, so that you’re in a position to entry the net and really get directed to the supply that it’s citing. NotebookLM was one thing we tried out, nevertheless it was not as profitable. It was a cool software that allowed us to summarize particular coverage paperwork right into a podcast, however the summaries have been typically fairly imprecise.
Crew NSC Power
Crew NSC Power was engaged on a Nationwide Safety Council drawback, “How can america generate adequate vitality to assist compute/AI within the subsequent 5 years?”
At the beginning of the category, the group started by utilizing ChatGPT to summarize their coverage papers and generate tailor-made interview questions, whereas Claude was used to synthesize analysis for background understanding. As ChatGPT often hallucinated info, by the top of the category they have been cross validating the summaries by way of Perplexity professional.
The group additionally used ChatGPT and Mermaid to prepare their ideas and decide who they wished to speak to. ChatGPT was used to generate code to place into the Mermaid flowchart organizer. Mermaid has its personal language, so ChatGPT was useful, so we didn’t must be taught all of the syntax for this language.
See the video of how Crew NSC Power used ChaptGPT and Mermaid right here:
In case you can’t see the video click on right here
Crew Alpha Technique
The Alpha Technique group was attempting to find whether or not the U.S. might use AI to create a whole-of-government decision-making manufacturing unit.
At the beginning of sophistication, Crew Alpha Technique used ChatGPT.40 for coverage doc evaluation and abstract, as nicely for stakeholder mapping. Nevertheless, they found going one after the other by means of the numerous numbers of articles was time consuming. So the group pivoted to utilizing Pocket book LM, for doc search and cross evaluation. See the video of how Crew Alpha Technique used Pocket book LM right here:
In case you can’t see the video click on right here
The opposite instruments the group used have been customized Gpts to construct stakeholder maps and diagrams and manage interview notes. There’s going to be all kinds of specialised Gpts. One which was actually useful, they stated, was a scholar GPT.
See the video of how Crew Alpha Technique used customized GPTs:
In case you can’t see the video click on right here
Like different groups, Alpha Technique used ChatGPT to summarize their interview notes and to create movement charts to stick into their weekly displays.
Crew Congress
The Congress group was exploring the query, “if the Division of Protection got financial devices of energy, which instruments could be only within the present techno-economic competitors with the Folks’s Republic of China?”
As different groups discovered, Crew Congress first used ChatGPT to extract key themes from a whole bunch of pages of readings every week and from press releases, articles, and laws. Additionally they used for mapping and diagramming to determine potential relationships between stakeholders, or to creatively counsel alternate visualizations.
When Crew Congress weren’t in a position to attain their sponsor within the preliminary two weeks of the category, very similar to Crew OSC, they used AI instruments to fake to be their sponsor, a member of the protection modernization caucus. As soon as they realized its utility, they continued to do mock interviews utilizing AI function play.
The group additionally used personalized fashions of ChatGPT however of their case discovered that this was restricted within the variety of paperwork they may add, as a result of that they had numerous content material. So that they used retrieval augmented technology, which takes in a person’s question, and matches it with related sources of their data base, and fed that again out because the output. See the video of how Crew Congress used retrieval augmented technology right here:
In case you can’t see the video click on right here
Crew NavalX
The NavalX group was studying how the U.S. Navy might broaden its capabilities in Intelligence, Surveillance, and Reconnaissance (ISR) operations on basic maritime site visitors.
Like all groups they used ChatGPT to summarize and extract from lengthy paperwork, organizing their interview notes, and defining technical phrases related to their challenge. On this video, observe their use of prompting to information ChatGPT to format their notes.
See the video of how Crew NavalX used tailor-made prompts for formatting interview notes right here:
In case you can’t see the video click on right here
Additionally they requested ChatGPT to function play a critic of our argument and resolution in order that we might discover the weaknesses. Additionally they started importing many interviews directly, and requested Claude to seek out themes or concepts in widespread that they could have missed on their very own.
Right here’s how the NavalX group used Perplexity for analysis.
In case you can’t see the video click on right here
Like different groups, the NavalX group found you’ll be able to customise ChatGPT by telling it the way you need it to behave.
In case you can’t see the video click on right here
One other stunning perception from the group is that you should utilize ChatGPT to inform you learn how to write higher prompts for itself.
In case you can’t see the video click on right here
In abstract, Crew NavalX used Claude to translate texts from Mandarin, and located that ChatGPT was the most effective for writing duties, Perplexity the most effective for analysis duties, Claude the most effective for studying duties, and pocket book LM was the most effective for summarization.
Classes Discovered
- Integrating AI into this class took a devoted teacher with a mission to create a brand new solution to educate utilizing AI instruments
- The end result was AI vastly enhanced and accelerated studying of all groups
- It acted as a useful collaborator
- Fusing AI with stakeholders interviews was particularly highly effective
- At the beginning of the category college students have been acquainted with a couple of of those AI instruments
- By the top of the category they have been fluent in lots of extra of them
- Most groups invented inventive use instances
- All Stanford courses we now educate – Hacking for Protection, Lean Launchpad, Entrepreneurship Inside Authorities – have AI built-in as a part of the course
- Subsequent 12 months’s AI instruments will probably be substantively higher
Filed beneath: Buyer Improvement, Gordian Knot Middle for Nationwide Safety Innovation, Nationwide Safety, Know-how Innovation and Nice Energy Competitors |
