I don’t know, the Google AI could be lying to me but it says the AQI readings on Weather Underground are real:
Weather Underground AQI readings are not generated by an AI model; they are actual, real-time measurements. [1]
Weather Underground aggregates this data directly from a few real-world, ground-level sources:
Official Government Monitors: Data is piped directly from the EPA/AirNow network in the US and the Copernicus Atmosphere Monitoring Service (CAMS) globally.
Community Sensor Networks: Weather Underground famously integrates readings from crowdsourced consumer air monitors, such as PurpleAir.
Proprietary Hardware: They also endorse their own commercial-grade home air sensors (the AQY 1). [1, 2, 3, 4, 5]
While AI is increasingly used in weather forecasting to predict how smoke or pollution might move, the base AQI numbers you see on Weather Underground are actual readings taken from these physical sensors. [1, 2, 3]
WeatherUnderground is an established science-based weather website. It is one of the industry standard sites for weather-related information (there are many others too). If you were to look at how many stations have temperature reports, and compare that to how many stations have air particles measurements, you would realize that there are only a limited number of locations that have particulate matter measurements. You have no information to make any claim that ai was used.
Thanks for the patience regarding my earlier post hinting at something new in tonearm land. I’ve been given the okay to talk about it, as the production details have been finalized. Here it is (note - this is a prototype, not the final version!):
This is a beta version of VPI’s new, upcoming true carbon fiber arm (shown in unipivot version, gimbal will also be available). This is the 12” version. I’ve used I think every version of VPI tonearm made, along with Jelco, Graham, Linn, SME, and more. Hey, good arms are out there, but in my own experience this new arm demonstrates how shockingly quiet an arm can be made to be. Dead silent, whip crack fast without being raw or analytical, open and cavernously deep low end, sweet high end, slightly wider soundstage. It does nothing wrong, that I can hear anyway. Tried it with a few different carts. They all play exceptionally better than my use of them in the past. More sound gets pulled out of the grooves than I’ve ever heard in the past.
I know there are arm makers who are critical of carbon fiber tubes. My use so far tells me that as all things in audio, a lot depends on the implementation. Is this arm “perfect”? I won’t say that. But IME it’s the best arm I’ve had the pleasure to use. It’s kind of like “you don’t really know what a tonearm is doing… until it isn’t doing it”.
Extra notes: the counterweight is being slightly redesigned in the production version. This beta version exhibits a rough finish on the tube because, well, it’s a prototype. Yes, using a new arm means all the carts have to be aligned and set up all over again, introducing changes that could be attributed to my observations. To this I can only say I’ve been around the block more than a few times with analog, and I use every quantitative tool in the book to make sure the measurable aspects are what they are for each of my carts. So cut me some slack in assuming I know how to do this stuff. And no, I don’t get anything from VPI for doing this kind of thing. It’s only because I’ve known Harry for so long, he knows I can set up a table properly, he knows he can trust my ears I guess, and I’m running a pretty darn good analog system that I occasionally get the opportunity to do stuff like this. I don’t even know what this arm will cost, meaning I don’t know if I’ll even be able to afford it!
I am familiar with WeatherUnderground, just suspicious of the number of claimed AQI sensors as reported. I am fully aware that many private/personal weather stations are utilized as inputs to WeatherUnderground. My “up north" neighbor has a WeatherUnderground weather station input as an example. They do not provide AQI inputs or measurements.
While not AI per se, it implies statistical modeling to provide the displayed information. Assuming the Google AI is correct. A questioning attitude never hurts, as the internet is devolving into something much different from that which we used to know. It now takes advantage of what it believes we want to be told, vice providing for hard truths.
I was somewhat tongue-in-cheek with the “I don’t know” disclaimer but I have read that some AI engines will “tailor” responses depending on what they think the questioner wants to hear. This mirrors your concerns I think.
Ron, the effective mass of this prototype is 16g, but the production unit will be14g. It’s a hollow carbon tube but is metallically shielded internally for EM. As for whether I’ll go with the unipivot (assuming I can afford these!), I should have the gimbal prototype soon so I’ll be able to compare. Wanna come to Florida and sweat your tuchas off?
BTW, I learned that carbon tubes can be made with a different modulus - essentially how springy or damped they are. VPI spent a ton of time experimenting with these to find the best modulus for their design goals.
I learned a lot about the different carbon layups, fibre orientation, tube modulus, resins, etc from my interest in carbon fiber bicycle frames and F1 chassis. It’s relatively limitless.
And no, I don’t get anything from VPI for doing this kind of thing. It’s only because I’ve known Harry for so long, he knows I can set up a table properly, he knows he can trust my ears I guess, and I’m running a pretty darn good analog system that I occasionally get the opportunity to do stuff like this. I don’t even know what this arm will cost, meaning I don’t know if I’ll even be able to afford it!
Wanna come to Florida and sweat your tuchas off?