The final experiment, woohoo we made it! I wanted to test Gemini to see if it could point out overlooked users. It did a great job, but I'm unsure how valuable this kind of insight is for this project. Anywho, we out ✌️

https://docs.google.com/document/d/1wk-cFectr4llESuFzroOXb33FR6NizeybAG3A6auSpg/edit?usp=sharing

#aiexperimentlog #1linkedin #medsuccess

I realized I never generated solutions based on my "how Might we" statement. Better now than never! Overall, it generated similar solutions to the ones I already had so... mid-success, somewhat insightful.

https://docs.google.com/document/d/1N2EuDBUKtw1je2uPeVGHIpuTEY-ABV2Z4xjpFgTo-Kc/edit?usp=sharing

#aiexperimentlog #1linkedin #medsuccess

Closing off for tonight, I had to test again AI's ability to fish out bad data. Used Perplexity, and I was so surprised by its success! Not only did it pick out my (purposefully) bad solutions from my good ones, but it also made suggestions on how to improve them. Pop off ig 💅

https://docs.google.com/document/d/1aE7AZ1DC9vauqaj16ka10LTz594lQPv09WXsZ5MIXLk/edit?usp=sharing

#aiexperimentlog #1linkedin #highsuccess

Today I asked NNG's ChatGPT for outlandish solutions that had nothing to do with my outlined pain points. It did a great job! This will set up my next experiment perfectly :)

https://docs.google.com/document/d/1GKALFtEwTYsNU1Iu8YGGGpdu3CmjSRfj8vA1KDxB_p4/edit?usp=sharing

#aiexperimentlog #1linkedin #highsuccess

Using Perplexity to generate solutions without any hints from me. Overall I did not gain much from this experiment, but it did validate some of my earlier ideas.

https://docs.google.com/document/d/19eTVqFjl0thvIjcgnVKVnMZyjXtj0qGVtsbda3p3drc/edit?usp=sharing

#aiexperimentlog #1linkedin #medsuccess

I created a solutions list and used Perplexity to analyze them. Nothing too complex, it did a nice job! I've been having too many successes though lately... I think soon I'll do another test including fake content of some kind 🤔 I want to see how smart these tools are

https://docs.google.com/document/d/17RX8dhTE8sGQcnth1as0akpXOQaqyqtgOv1xxgu1U_4/edit?usp=sharing

#aiexperimentlog #1linkedin #highsuccess

Kicking off the final phase woohoo! To start, I conducted the 5 whys and asked the NNG ChatGPT to analyze my responses. It helped to improve my answers and keep me on track!

https://docs.google.com/document/d/1keKbNYph95raP4CYLFkSZu8Fwg4E0jlS1I1GTKVpbvA/edit?usp=sharing

#aiexperimentlog #1linkedin #highsuccess

I was inspired by Friday's class to generate a HMW question. I've struggled with ChatGPT in the past, but I used the NNG Chat for the first time, and I felt this experiment was much better! I landed on a question that feels balanced and relevant to my group's problem.

https://docs.google.com/document/d/1aru1qj2kycERZ2BBMyeV67fysyxZWgl3etsTAID0sCk/edit?usp=sharing

#aiexperimentlog #1linkedin #highsuccess

Tested Notebook LLM to see if anything was missing from my user journey map. It did provide some good insights which led to minor changes. Overall it did a good job 👍

https://docs.google.com/document/d/1-qaVsBniF5pSvE9ofQsL9t59becBV3TILg2rNOumomQ/edit?usp=sharing

#aiexperimentlog #1linkedin #medsuccess

Been enjoying Claude lately! This time I used it to generate information for a user journey map. I wasn't expecting it to generate the map itself too! It did a nice job with this, I was impressed :)

https://docs.google.com/document/d/1JQcLhppcOBCZuHZTo65gQyxpRDdm1edGp1ufyYM7qYE/edit?usp=sharing

#aiexperimentlog #1linkedin #highsuccess