Strategia ja liiketoiminta, tekoäly

The 5-months AI Builders Lab recap: This is what we did and this is what I learned

By far the best AI course I ever had.

So last week we wrapped up our five months long AI Journalism Lab: Builders at Craig Newmark Graduate School of Journalism at CUNY (City University of New York), in partnership with Nordic AI Journalism.

Shoutout especially to Aldana Vales and Marie Gilot for organizing this so professionally, and of course to the whole amazing cohort of 24 people and our coaches and speakers such as Nikita Roy, Lukas Nielsen, Kasper Lindskow, Joe Amtidis, Upasna Gautam, Elite Truong and others.

The curriculum of the program consisted of these themes:

  • Designing AI products for newsrooms
  • Prototyping with No-Code Tools
  • (Un)Structured data, Semantic Search & RAGs
  • AI Agents & Autonomous Workflows
  • Measurement & Ethics in the Product Lifecycle
  • Launching and Sustaining Products in an AI Future

Each one of us built a final project during these five months, from which you will hear a bit more later on. Mine unfortunately is not public information since I’ve just changed jobs but what I can tell is that it is a python-based tool that has to do with videos, it was built with Google Antigravity, and I’m really happy with how it turned out.

What did I learn?

A whole lot of new stuff, starting from how to actually build a good PRD, product requirement document for AI to use. I will also definitely use several of the newsroom spesific frameworks we came across on the course, like one that helps evaluate and prioritize AI products or features from five different perspectives simultaneously. All in all, I’d say I learned a lot from…

  • How to identify whether a problem is genuinely an AI problem
  • How to prioritize AI development by value and feasibility
  • How to build prototypes fast
  • How the underlying data and AI infrastructure works and where it typically breaks
  • How to evaluate an AI product’s value across multiple dimensions simultaneously
  • How to deal with the fact that building the tools is not usually the hardest part (but scaling)

So why was this the best AI course I ever had?

It was both strategic and concrete. Spesific newsroom context, really hands on. Excitement but not overhype. Realism.

What’s also nice is that our cohort will continue to have discussions in our joint Slack group that will remain after the course as well.

So, if you have the possibility to attend some of CUNY’s offerings, I highly recommend. Check it out here.

Normaali

Jätä kommentti