The Alexa feature "do not send voice recordings" you enabled no longer available

Post by pc86 » Mon, Mar 17, 2025, 12:58 PM UTC

latexr wrote: post #286

pc86 wrote: post #275

Earlier quoted context omitted.

I have to imagine you cherry-picked maternal death rate (which is anomalously high in the US for reasons that even experts in maternal death cannot explain) in bad faith because the all-cause mortality rate in the US is 1,044 per 100k compared to 1,412 per 100k in Greece - 35% higher[0]. I guess it's not so clear. [0] https://worldpopulationreview.com/country-rankings/death-rat...

> which is anomalously high in the US for reasons that even experts in maternal death cannot explain Doesn’t seem like a mystery that a country who is so “pro-life” they’d rather let women die than properly treat miscarriages has a high maternal death rate. https://www.texastribune.org/2024/11/27/texas-abortion-death...

Post by AmVess » Mon, Mar 17, 2025, 12:59 PM UTC

ropejumper wrote: post #273

bko wrote: post #264

Earlier quoted context omitted.

Take pretty much any quality of life thing. Just as one example Mississippi: The average home size 2,065 square feet United Kingdom: The average home size is 818 square feet No, you cannot "eat GDP", but its a useful proxy as it correlates very highly to a lot of things we do care about

This requires a justification for why home size is a good indicator of quality of life, or for why it's more important than other indicators.

Post by pzo » Mon, Mar 17, 2025, 1:00 PM UTC

Dylan16807 wrote: post #63

pzo wrote: post #52

Earlier quoted context omitted.

LLM can take audio as input, e.g. Gemini 2.0 Flash can take audio as input and is very fast and cheap (costs 25 audio tokens for a second, and prices $0.1 for 1 million input tokens). But I get your point that e.g. whisper is also very good and can run fast locally on edge devices. We also have recently phi4-multimodal LLM that is 5.6B params and is good enough to do inference locally on beefy smartphones. Problem is…

Is the specific component that takes the audio in gemini an LLM? If so, how does audio data become tokens? But my point is not just about the ease of running whisper, it's that the device is already doing transcription. Even if it's not optimal, it works.

Post by HelloMcFly » Mon, Mar 17, 2025, 1:01 PM UTC

delecti wrote: post #254

etrautmann wrote: post #243

Earlier quoted context omitted.

But why do we need these in the first place? I have a free Apple home that I can’t find much of a use for aside from being a speaker. Is it that hard to get the weather or turn lights off from your phone? I definitely don’t want to order products from an audio interface. What else am I missing?

What if you aren't currently holding your phone, with it unlocked and the (probably slow to load) app open? What if you don't have switched outlets, and want to turn off lights either a) when the room is still dark, or b) when you're leaving the room? Voice assistants are not a revolution in home living, but lets not go too far in the other direction and ignore that they do still have useful features.

Post by HideousKojima » Mon, Mar 17, 2025, 1:03 PM UTC

xandrius wrote: post #291

bko wrote: post #264

Earlier quoted context omitted.

Take pretty much any quality of life thing. Just as one example Mississippi: The average home size 2,065 square feet United Kingdom: The average home size is 818 square feet No, you cannot "eat GDP", but its a useful proxy as it correlates very highly to a lot of things we do care about

Population density Mississipi: 24.5/km2 UK: 279/km2 Maybe, only maybe, that has an effect on house sizes?