After tracking the recent fires in LA & Hawaii, the massive blackouts in Europe, and other recent destructive flood/wind events, I realized how unprepared we are as a population for these types of severe events (myself included). After chatting with a lot of folks, I realized people are generally aware of these issues - but aren't sure where to start, what specific steps to take, or how to finance the whole thing.
My friend and I wanted to build a tool that does three things:
1. Give an immediate, address-level risk profile for natural hazards (starting with fire/flood) and crime vulnerability.
2. Generate a practical 'improvements' checklist to "harden" that specific property.
3. Help surface rebates, credits, and other free services - to make it easier for people to afford these improvements.
Even as I’ve spent the last year researching natural hazards, one fact still haunts me: Windborne wildfire embers (firebrands) can travel up to 5 miles ahead of the main fire front.
Most homeowners look at their immediate neighborhood and think, "I’m not near the trees, I’m safe." But the data shows that's a fallacy. You aren't just at risk from your backyard; you’re potentially at risk from a 5-mile radius of backyards.
In the context of home safety, a wildfire might be a Black Swan to the average person, but a Dragon King to Didier Sornette who might see the fuel load, the wind patterns, and the 5-mile ember-throw potential.
This is an early version so your thoughts would be helpful:
- Any bugs or unclear features?
- Thoughts on how to show this more clearly to users.
- If you were looking for a new house (or improve an existing one), would using this tool help you feel safer?
Personally-speaking, there aren't that many different consumer technologies or tech applications over the past 2+ decades that truly gave me an 'aha' moment. The Palm Pilot, Google Maps on an early Android phone come to mind.
When I began playing around with LLMs, I had my initial aha moment. It was far above either of those for me. So, I think that collective aha moment we've been having the last few years (still) drives a lot of the excitement/hype.
If you view 'hype around [x]' as essentially a probability ranking problem, that is, whatever is most likely to give you your next aha moment generates the most hype, at any point in time. There's a decay element, too, if the reality doesn't match the expectations, then other technologies are viewed as more likely to produce that next big aha moment.
But for now I think this characterizes AI today more than other technologies.
You made me curious. This reads almost like reasoning mode. I put this prompt in myself and got a much more structured response. Was this question asked at the end of a long discussion/context window - or was it the only input?
Yep. And I think many are too focused on just the pure ability to build models now, when in reality a model is at its core a set of assumptions. Using models you don't understand, didn't build, and barely reviewed amplifies the magnitude of any assumption error, imo.
My friend and I wanted to build a tool that does three things: 1. Give an immediate, address-level risk profile for natural hazards (starting with fire/flood) and crime vulnerability. 2. Generate a practical 'improvements' checklist to "harden" that specific property. 3. Help surface rebates, credits, and other free services - to make it easier for people to afford these improvements.
Even as I’ve spent the last year researching natural hazards, one fact still haunts me: Windborne wildfire embers (firebrands) can travel up to 5 miles ahead of the main fire front.
Most homeowners look at their immediate neighborhood and think, "I’m not near the trees, I’m safe." But the data shows that's a fallacy. You aren't just at risk from your backyard; you’re potentially at risk from a 5-mile radius of backyards.
In the context of home safety, a wildfire might be a Black Swan to the average person, but a Dragon King to Didier Sornette who might see the fuel load, the wind patterns, and the 5-mile ember-throw potential.
This is an early version so your thoughts would be helpful:
- Any bugs or unclear features? - Thoughts on how to show this more clearly to users. - If you were looking for a new house (or improve an existing one), would using this tool help you feel safer?
Thanks!