6 ms·

Serious question: is it typical to describe client-side computing as "on the edge"? I thought running something on the edge referred to running it in close net

Serious question: is it typical to describe client-side computing as "on the edge"? I thought running something on the edge referred to running it in close network proximity to the user, rather than users having control and running things themselves.

capableweb 3y ago
Yes, "edge computing" can refer to both computing done as close to the user as possible geographically, or even on the device itself. If someone says "I wanna do edge computing" it's not clear enough to know if they just want to have servers they control as close to the user as possible, or do the computing on the device itself. I think Apple would say "edge computing" is on the actual device while CloudFlare would say "edge computing" is on their infrastructure, but distributed to be physically closer to the end user.
- iamerroragent 3y ago
  I guess I've been out of the loop for a bit and didn't realize that "edge computing" became a term since cloud computing took off. It is kind of cyclical then is not? By that I mean computers used to be shared and to log into it through a terminal. Then the PC came around. Then about 15 years ago Cloud computing became the rage (really an extension or more sophisticated system than the first time shared computers) Now we're back to local computing. I even see more self hosting and moving away from cloud due to costs. All that rant is to say is it's interesting. Side note, getting this AI to be localized as much as possible I imagine will be really useful in the medical industry because it helps alleviate HIPAA requirements.
  - nordsieck 3y ago
    > It is kind of cyclical then is not? > By that I mean computers used to be shared and to log into it through a terminal. > Then the PC came around. > Then about 15 years ago Cloud computing became the rage (really an extension or more sophisticated system than the first time shared computers) There's a really neat article called "The Eternal Mainframe"[1] that you might be interested. It explores this idea in greater depth. --- 1. http://www.winestockwebdesign.com/Essays/Eternal_Mainframe.html http://www.winestockwebdesign.com/Essays/Eternal_Mainframe.h...
    - iamerroragent 3y ago
      Thanks, that was an interesting read! I wonder if the author's perspective has changed with regards to freedom to compute. Social Media is often used as an example of privacy invasion though I've failed to see why concerns over Facebook handling your private data is worrying when they don't have a product you need to have. Email on the other hand, is pretty much a necessity today so privacy concerns are vital there imo. Of course you can host your own server whereas you can't host your own Facebook.
      - macintux 3y ago
        > I've failed to see why concerns over Facebook handling your private data is worrying when they don't have a product you need to have. I have at least two concerns. 1) Viscerally, it would be intensely creepy if someone started following me everywhere. Why would I want a company doing the same? 2) We are in an era of diminishing freedoms, and it's impossible to prove that data Facebook has on us can't be turned against us in the court system in the future. Women's menstrual cycles are being weaponized, for Pete's sake.
        iamerroragent 3y ago
        You're completely correct. What I wanted to articulate though was that Facebook isn't delivering an essential service. The consumer still has a choice not to use it. So if it as a company is known to abuse the information you give it then don't do business with that company. Email, phone, and even your personal address you don't have a lot choice around not using. And that stuff also isn't really protected without encryption. (excluding mail by postal service, mostly). Basically if you don't want communication to be intercepted and potentially used against you then do it in person inside of faraday cage. I'm mostly joking here.
        lovemenot 3y ago
        I have never had a Facebook account. I believe Meta would have a shadow-account for me, based on browser fingerprint, photos tagged by their users and such. At some point, one's own choices become moot.
wsgeorge 3y ago
I believe this has been extended to mean "on device", which is interesting. See Gerganov's article on Github [0]. I wrote about this here [1] where I made a contrast between the core and the edge. I think the term maps well to this meaning. What I find more interesting is that in the classic "close network proximity", some parts of the world may not have benefited as much from that trend since the closest nodes of a global delivery network could be several countries away. [0] https://github.com/ggerganov/llama.cpp/discussions/205 https://github.com/ggerganov/llama.cpp/discussions/205 [1] https://medium.com/sort-of-like-a-tech-diary/consumer-ai-is-ripe-for-centralization-f3251dfa460e https://medium.com/sort-of-like-a-tech-diary/consumer-ai-is-...
- TeMPOraL 3y ago
  > I believe this has been extended to mean "on device", which is interesting. I don't like the connotations this carries. This is almost openly talking about reaching all the way into peoples' hardware to run your software, for your benefit, on them, without their knowledge, consent or control...
  - wsgeorge 3y ago
    I see. Hadn't considered this. Yes, I see how that might be a concern. What I think is important in this AI Spring is that we make it possible for people to run their own models on their own hardware too, without having to submit anything to a large, centralised model for inference.
dragonwriter 3y ago
> Serious question: is it typical to describe client-side computing as “on the edge”? Somewhat; its consistent with, e.g., Google’s “Edge TPU” designation for its client-side neural processors. > I thought running something on the edge referred to running it in close network proximity to the user Typically, but on the client device is the limit-case of “close network proximity to the user”, so the use is consistent.
layer8 3y ago
“Edge computing” arguably implies there’s a network you are connected t, that you’re on the edge of, so I wouldn’t apply the term to applications that can function completely offline. With edge computing there’s usually still a notion of having some sort of internet integration, like IoT devices.
dannyobrien 3y ago
I've used "edge" in this context for around 15 years[1], and I've always intended it to mean "at the edge of the network", which can include being on the other side of the world to a user. [1] from https://www.oblomovka.com/wp/2007/08/ https://www.oblomovka.com/wp/2007/08/ at least
aargh_aargh 3y ago
Because of the ambiguity of the term "on the edge" that is used to refer to both close network proximity and the device closest to the user, as evidenced by this thread, I would suggest to use a new term, at least in the context of A.I. The AI running on the device closest to the user should be called a "terminator".

capableweb 3y ago

Yes, "edge computing" can refer to both computing done as close to the user as possible geographically, or even on the device itself. If someone says "I wanna do edge computing" it's not clear enough to know if they just want to have servers they control as close to the user as possible, or do the computing on the device itself. I think Apple would say "edge computing" is on the actual device while CloudFlare would say "edge computing" is on their infrastructure, but distributed to be physically closer to the end user.

iamerroragent 3y ago
I guess I've been out of the loop for a bit and didn't realize that "edge computing" became a term since cloud computing took off. It is kind of cyclical then is not? By that I mean computers used to be shared and to log into it through a terminal. Then the PC came around. Then about 15 years ago Cloud computing became the rage (really an extension or more sophisticated system than the first time shared computers) Now we're back to local computing. I even see more self hosting and moving away from cloud due to costs. All that rant is to say is it's interesting. Side note, getting this AI to be localized as much as possible I imagine will be really useful in the medical industry because it helps alleviate HIPAA requirements.
- nordsieck 3y ago
  > It is kind of cyclical then is not? > By that I mean computers used to be shared and to log into it through a terminal. > Then the PC came around. > Then about 15 years ago Cloud computing became the rage (really an extension or more sophisticated system than the first time shared computers) There's a really neat article called "The Eternal Mainframe"[1] that you might be interested. It explores this idea in greater depth. --- 1. http://www.winestockwebdesign.com/Essays/Eternal_Mainframe.html http://www.winestockwebdesign.com/Essays/Eternal_Mainframe.h...
  - iamerroragent 3y ago
    Thanks, that was an interesting read! I wonder if the author's perspective has changed with regards to freedom to compute. Social Media is often used as an example of privacy invasion though I've failed to see why concerns over Facebook handling your private data is worrying when they don't have a product you need to have. Email on the other hand, is pretty much a necessity today so privacy concerns are vital there imo. Of course you can host your own server whereas you can't host your own Facebook.
    - macintux 3y ago
      > I've failed to see why concerns over Facebook handling your private data is worrying when they don't have a product you need to have. I have at least two concerns. 1) Viscerally, it would be intensely creepy if someone started following me everywhere. Why would I want a company doing the same? 2) We are in an era of diminishing freedoms, and it's impossible to prove that data Facebook has on us can't be turned against us in the court system in the future. Women's menstrual cycles are being weaponized, for Pete's sake.
      - iamerroragent 3y ago
        You're completely correct. What I wanted to articulate though was that Facebook isn't delivering an essential service. The consumer still has a choice not to use it. So if it as a company is known to abuse the information you give it then don't do business with that company. Email, phone, and even your personal address you don't have a lot choice around not using. And that stuff also isn't really protected without encryption. (excluding mail by postal service, mostly). Basically if you don't want communication to be intercepted and potentially used against you then do it in person inside of faraday cage. I'm mostly joking here.
        lovemenot 3y ago
        I have never had a Facebook account. I believe Meta would have a shadow-account for me, based on browser fingerprint, photos tagged by their users and such. At some point, one's own choices become moot.

wsgeorge 3y ago

I believe this has been extended to mean "on device", which is interesting. See Gerganov's article on Github [0]. I wrote about this here [1] where I made a contrast between the core and the edge. I think the term maps well to this meaning. What I find more interesting is that in the classic "close network proximity", some parts of the world may not have benefited as much from that trend since the closest nodes of a global delivery network could be several countries away. [0] https://github.com/ggerganov/llama.cpp/discussions/205 https://github.com/ggerganov/llama.cpp/discussions/205 [1] https://medium.com/sort-of-like-a-tech-diary/consumer-ai-is-ripe-for-centralization-f3251dfa460e https://medium.com/sort-of-like-a-tech-diary/consumer-ai-is-...

TeMPOraL 3y ago
> I believe this has been extended to mean "on device", which is interesting. I don't like the connotations this carries. This is almost openly talking about reaching all the way into peoples' hardware to run your software, for your benefit, on them, without their knowledge, consent or control...
- wsgeorge 3y ago
  I see. Hadn't considered this. Yes, I see how that might be a concern. What I think is important in this AI Spring is that we make it possible for people to run their own models on their own hardware too, without having to submit anything to a large, centralised model for inference.

dragonwriter 3y ago

> Serious question: is it typical to describe client-side computing as “on the edge”? Somewhat; its consistent with, e.g., Google’s “Edge TPU” designation for its client-side neural processors. > I thought running something on the edge referred to running it in close network proximity to the user Typically, but on the client device is the limit-case of “close network proximity to the user”, so the use is consistent.

layer8 3y ago

“Edge computing” arguably implies there’s a network you are connected t, that you’re on the edge of, so I wouldn’t apply the term to applications that can function completely offline. With edge computing there’s usually still a notion of having some sort of internet integration, like IoT devices.

dannyobrien 3y ago

I've used "edge" in this context for around 15 years[1], and I've always intended it to mean "at the edge of the network", which can include being on the other side of the world to a user. [1] from https://www.oblomovka.com/wp/2007/08/ https://www.oblomovka.com/wp/2007/08/ at least

aargh_aargh 3y ago

Because of the ambiguity of the term "on the edge" that is used to refer to both close network proximity and the device closest to the user, as evidenced by this thread, I would suggest to use a new term, at least in the context of A.I. The AI running on the device closest to the user should be called a "terminator".