Pete Lehmann PRO
BigBallinCrackerAss
·
AI & ML interests
I am a straight up user of AI, I don't have the knowledge or skills to be part of the community actually developing the AI models themselves. That said, I use ChatGPT and Gemini primarily, depending on what task I'm trying to accomplish. ChatGPT is my go to for working through personal issues using the chat bot, but also I have developed a few focused GPTs to use at my job as an airport manager. I front loaded it with the specific FAA regulation, policies, and guidance related to the operations of public use airports, as well as the FAA's airport grant programs and the incredibly complex requirements for airports who accept those grant funds. I also tend to use it for the everyday simple tasks like writing a formal letter, questions about random topics, and all that kind of stuff.
I never used Gemini for much until something happened an that model improved drastically! It's my go to AI for anything involving my Google account features: Gmail, calendar, YouTube, etc. I absolutely LOVE NotebookLM and use it regularly for reasons that I'm pretty sure the developers weren't thinking of during development. Sure, I let it help me with presentations and the infographics are super cool.
The way I deal with issues or conflicts in my life is to write..... a LOT! In fact I have a folder I've been keeping for about a decade and a half named "Nonsensicals" which is chock full of these brain dumps, and they are numbering somewhere between 550-600 unique documents. What I do with NotebookLM is I upload one of these documents which I just spent a couple hours pouring out of my brain which is generally about something I am still very emotional about or I feel is not yet resolved in my mind, sometimes 2 documents, but only if they are directly related to the same topic. Then I have NotebookLM create one of the longer podcasts which is always a man and a woman chatting about what I wrote.
Well I quickly realized that there discussions about the issues I had just written about were often looking at my issue from a completely different perspective than what I wrote. In almost every instance the podcast created from my own writings ended up providing me with a new level of insight that I would not have otherwise recognized. Absolutely LOVE that tool and it just keeps getting better.
Gemini is also leaps and bounds ahead of ChatGPT with any type of photo or video work, editing, text prompt creation, animating. etc. But where I have found that it excels is in the creation of custom applications. The most recent success was actually an interesting exercise in getting frustrated at it's inability to create what I needed, so I waited 6 months, went back to the same project, and it knocked it out of park! As an airport manager I have to ensure that all of the asphalt surface movement areas on my airport (runway, taxiway, taxilanes, aprons) are inspected regularly for damage or wear which could cause an aircraft to lose control while taxiing, taking off, or landing, which could lead to an accident or incident. There was really no established method or tool to record any issues I might find, and if an issue was found and it was noted, there was no accurate, consistent, or intuitive way to describe the location of the issue on the surface areas to ensure I can find the issue again easily. So Gemini and I pulled an all nighter and took what we had started months earlier, the Eastport Municipal Airport Surface Movement Area Condition & Maintenance Reporting System, and created in the end a full screen graphic of the airport with all the asphalt, hangars, fuel farm, and terminal building placed where appropriate, and a sidebar on the right where a summary of any active records are displayed. To create a new record to log, for example a damaged taxiway edge, I simply move my mouse curser over the approximate location of the damage on the graphic, double click, and the new record dialogue box pops up for me to fill in the details. Once I've entered in all the information I have, I select one of the color coded severity tiers, and click save. The dialogue box disappears, the new record summary appears on the right sidebar, and a small circle icon appears where I double clicked colored based on the severity. There's a ton more features and UI awesomeness, but I'm done typing for now.
Recent Activity
new activity about 2 months ago
multimodalart/qwen-image-multiple-angles-3d-camera:Just Have to Say Dang! liked a Space about 2 months ago
multimodalart/qwen-image-multiple-angles-3d-cameraOrganizations
None yet