Make up your mind Google AI. Is sound faster in air that is less dense or more dense?

Honestly, there is so much wrong in the AI answers that it’s hard to know where to start, but the direct contradiction of itself seems like a good start.

14 points

Its all just garbage that gets in the way. My W11 systems was randomly bogging down during documents, excel and powepoint. it was the AI service hogging resources, (c packaged with Office.) Easy fix, just delete the AI executables in a folder, but a product update will probably bring the back.

permalink
report
reply
5 points

And the product updates will happen without you asking them to! And if you disable them, seemingly unrelated windows updates will helpfully fix your mistake.

permalink
report
parent
reply
17 points

This seems like a difficult thing to get right. To me it would intuitively seem like air transmits sound easier than e.g. water or steel since there is less to dampen the waves. But that’s just wrong. You shouldn’t trust intuition when it comes to physics, even if you are a physicist.

permalink
report
reply
5 points

I’m not a physicist and I’m taking a stab without looking it up.

Is there an index of refraction or something so that if you transmit through air than water than air you would think that sound is “dampened”?

I’m having a hard time with just an object though. Is the speed of sound a constant? I’ve never heard it defined with respect to to air

permalink
report
parent
reply
10 points
*

I am a physicist, for context. Please just use a table for the values in air.

What do you mean with index of refraction? For light this refers to the speed of light in the medium. In this sense you can define a index of refraction for sound, but would you want to? It has very little to do with dampening (dampening is usually wave length depended so they are usually proportional).

The speed of sound in air is ca. 300m/s, in water 1500m/s. so their relative index of refraction are 5. This implies rather difficult transition of sounds between medias since most sounds are going to be reflected. Refer to frustrated reflection.

The physics of why denser air is claimed to have a slower speed of sound is not clear to me, but I suspect there is some bullshit going on since the question is not sensical. You can’t double the density of a gas without changing other parameters like Temperature or Pressure. Refer to the ideal gas law.

permalink
report
parent
reply
4 points

There’s an easy formula for ideal gases: c = sqrt( gamma * R * T ) = sqrt( gamma * P / rho ). [Express ideal gas law as P = rho * R * T using a gas constant tailored to your species].

So in isobaric (equal pressure) conditions, there is an inverse relationship between speed of sound and density.

But the atmosphere is not isobaric, especially not on its vertical axis. For the first layer of atmosphere, the vertical profile can be roughly characterized by a linear drop in temperature from sea level to 11 km altitude. In this region the speed of sound is therefore also dropping linearly, but the air is also getting less dense.

Source: programmed air data software for aircraft.

permalink
report
parent
reply
1 point

Hm, okay. So sound /does/ travel at different speeds in different mediums. Haven’t heard that before.

I guess the refraction (defined the way you assumed) has applications with noise reduction in, say, a building.

As sound appears to travel faster through liquids than gas, I would imagine the answer is the particles don’t have to travel as far to transmit energy if the particles are closer?

permalink
report
parent
reply
2 points

Sounds like maybe you want acoustic impedance ? Just like optical index mismatch, or electrical impedance mismatch, you get reflections at discontinuities. Neat stuff!

Not exactly sure what you mean by air-water-air “dampening,” but my suspicion is that you’re referring to sound being reflected at each interface, so the transmission is reduced. Antireflective coating, index matching, impedance matching are all rich topics in physics and electronics!

permalink
report
parent
reply
2 points

Yes the transmission reduction you need mentioned. Thank you

permalink
report
parent
reply
9 points

I have found similar contradictions for biology searches. AI in many ways is just a glorified search engine, and it makes mistakes based on what’s available to it.

permalink
report
reply
6 points

At least a traditional search engine doesn’t torch an couple acres of rainforest just to give you wrong information.

permalink
report
parent
reply
2 points

AI is like an accountant who answers “What do you want it to be?” when you ask what 2 + 2 is.

permalink
report
reply
4 points

I don’t understand why people are complaining about these shitty features instead of just turning them off.

permalink
report
reply
27 points

Because it can have negative impacts on the lives of people who aren’t savvy enough to double check the info. There’s already enough misinfo on the internet. We don’t need a (formerly) trusted source spreading more.

permalink
report
parent
reply
11 points

Not everyone is savvy enough to turn it off, for one. The average person isn’t even going to think about turning it off. That means a lot of people are now being fed a top search result that is the wrong info half of the time. Not just the wrong webpage, but actually the wrong information.

For another thing, it shouldn’t be on by default if it’s so bad. If this was a traditional bug giving you incorrect search results half the time, it wouldn’t be released. But because of this AI race that’s happening, google is willing to release this massive bug live, and on by default. We should be complaining about it!

I also think part of the problem is that it seems really useful. At first glance it seems like it has quickly and succinctly summarized the information that is deep inside other web pages, and presented the answer to specifically what i was looking for (quite confidently, at that.) It’s very easy to fall into a trap of trusting the information told to you.

permalink
report
parent
reply
2 points

Yeah I was addressing the audience here on Lemmy though. I get thinking the feature sucks but you can turn it off, which I did a while ago. I think it would make a lot more sense to complain about this in a setting like reddit where you’re not preaching to the choir so directly.

On Lemmy, I feel like 90% of users can build an app from source and debug dependency issues to make it happen. So it’s just odd to me that I still see this getting beat to death here.

permalink
report
parent
reply
2 points

Just because you can turn it off for yourself doesn’t mean the problem no longer exists.

permalink
report
parent
reply
4 points

“I neeeeed them for wooorrrrrrrk”

I dunno. The hypetrain might be the biggest we’ve seen in our lifetimes, relative to the actual impact of the, y’know, thing. That’s Trillions of Quattloos worth of hype out there pumping the lies of what it can do to people who can’t remember how to clear the cache.

permalink
report
parent
reply
2 points

please show me the off button

permalink
report
parent
reply
2 points

In a browser, this will help: https://tenbluelinks.org/

If in the mobile app, look in settings and there are settings related to AI and “labs” that you can turn off.

permalink
report
parent
reply

Artificial Ignorance

!ArtificialIgnorance@lemmy.ca

Create post

In this community we share the best (worst?) examples of Artificial “Intelligence” being completely moronic. Did an AI give you the totally wrong answer and then in the same sentence contradict itself? Did it misquote a Wikipedia article with the exact wrong answer? Maybe it completely misinterpreted your image prompt and “created” something ridiculous.

Post your screenshots here, ideally showing the prompt and the epic stupidity.

Let’s keep it light and fun, and embarrass the hell out of these Artificial Ignoramuses.

All languages welcome, but an English explanation would be appreciated to keep a common method of communication. Maybe use AI to do the translation for you…

Community stats

  • 312

    Monthly active users

  • 3

    Posts

  • 58

    Comments

Community moderators