It was windy today in Annapolis and we lost power for a bit (see below for the weather warning). Note: these warnings may not meet plain language standards but they are not intended to. The warning covers a large geographical area and they are written for multiple persona: emergency preparedness managers, local authorities and the general public. LLM's can act as a great assistant here e.g.: "I am a INSERT_PERSONAL and I have just received a weather warning INSERT_WEATHER_WARNING. Can you summarize this and how should I prepare?"
Simon Wilson's LLM CLI is really handy for quick useage of an LLM.
curl "https://forecast.weather.gov/showsigwx.php?warnzone=MDZ014&warncounty=MDC003&firewxzone=MDZ014&local_place1=Annapolis" | llm -s "I am a resident of Annapolis and I have just received a weather warning. Can you summarize this and how should I prepare?"
The Department of Information Technology, State of Maryland maintains a live dataset of power outages thanks to the Maryland’s Open Data Portal . We can see how many customers are without power in Anne Arundel county by doing:
curl -s "https://opendata.maryland.gov/api/views/uxq4-6wxf/rows.csv" \
| awk -F, 'NR==1 { print; next } $2=="Anne Arundel" { latest = $0 } END { print latest }' | llm -s "How many customers are without power in Anne Arundel county and what percentage of customers are without power?"