6 ms·

We always assumed they wouldn't spill their beans to third parties. I guess that was assuming too much.

bbctol 10y ago
I think we just envisioned a highly anthropomorphized AI: essentially, a very smart and entirely obedient person to serve as the perfect aide. The Star Trek dream emerged before computing technology was very far advanced, and well before the idea of constantly mobile wireless communication. We thought our AIs would be small and physical, easy for a single person to entirely own and unable to remember more details than a human; instead we got unfamiliar algorithms run on machines far away.
- dragonwriter 10y ago
  > The Star Trek dream emerged before computing technology was very far advanced, and well before the idea of constantly mobile wireless communication. We thought our AIs would be small and physical, easy for a single person to entirely own and unable to remember more details than a human; instead we got unfamiliar algorithms run on machines far away. Which is kind of funny, because even if it might be accessed by personal mobile devices, the Star Trek "library computer" AI was never "small and physical, easy for a single person to entirely own and unable to remember more details than a human", it was an aspect of a large server (or networked cluster, the actual architecture is somewhat vague) that was part of a capital ship or base, had access in the server/cluster to a library of very nearly all generally available knowledge and extensive personal information about both its users and about people with little direct connection (and could reach out across a galactic network to access additional remote information sources to handle requests). "Unfamiliar algorithms run on machines far away" is much more like the source of the "Star Trek dream" than "small and physical, easy for a single person to entirely own and unable to remember more details than a human".
  - bbctol 10y ago
    True! I suppose I was thinking of earlier space-age fantasy more than TNG. It was rather odd how easily people could look up details on crewmembers aboard the Enterprise, but I suppose the situation's different on an organized ship.
    - dragonwriter 10y ago
      > It was rather odd how easily people could look up details on crewmembers aboard the Enterprise, but I suppose the situation's different on an organized ship. Not just crewmembers: TNG showed some of the broader implications (both useful and creepy) of the convenience-oriented panopticon, e.g., when Geordi used the Enterprise library computer to construct a simulation from data (including personality profiles) of Dr. Leah Brahms, who later meets him (and encounters the simulation.) (S306 "Booby Trap", S416 "Galaxy's Child")
      - VLM 10y ago
        That was a substantial plot hole where she had no idea someone was simulating her. I suspect a huge part of the panopticon culture would be / is being informed that you're being peeped at. 99% of the time someone asking "Computer locate commander Riker" involved commander Riker knowing all about who's looking for him and why and having a substantial conversation with the requester. I don't recall any plot along the lines of Deanna getting jealous and spamming the computer all night asking where Riker is and he better not be in that cute ensign's bedroom... Because it seems logical the computer would inform him each time and he would eventually tire of the interruption and nature would take its course, WRT his relationship with Deanna. A better analogy would be technically I could walk up to the company president's office and stalk him, but culturally that is so not going to fly and I would have a lot of explaining to do. Merely using the computer instead of walking there in flesh isn't a major cultural shift. However, every single "holodeck creeper" plot line involved the simulated attractive real world woman not knowing she's being simulated until the plot reached maximum spaghetti spilling cringe, which is one of the few Trek panopticon situations where people being spied on did NOT know they were being spied on, which seems very un-trek, although it made for some entertaining stories. An alternative interpretation is I believe over the course of the series every attractive woman on the ship was simulated on the holodeck at least once by at least one lonely guy, and its possible that culturally they just got used to it, although I find that unlikely. People do get conditioned to become used to the weirdest things, so its not out of the realm of possibility. Possibly a culture of what happens in vegas stays in vegas develops and its just the sexism of the TV show that they never showed the women turning the tables on their fellow male crewmen on the holodeck.
  - Spoom 10y ago
    I think Star Trek was very prescient here as well: Security is terrible in the Star Trek universe. Nearly every antagonist is able to break into the computer within about a minute.
- gvurrdon 10y ago
  Isn't it the case that Apple's attempts at similar AI assistance store much of the data on the user's iPhone?
mattmanser 10y ago
It really is the big problem at the moment with the cutting edge of AI. ML relies on large data-sets and if anyone tried to release a personal device it simply wouldn't even work, let alone compete with the mass surveillance google/ms/amazon are bringing to bear. Unless the state-of-the-art in AI suddenly morphs, we seem to be stuck between giving up our privacy or having vaguely intelligent AI. I personally fall heavily on the privacy side of stuff, but I can see the intellectual and commercial appeal of pretending it doesn't matter in order to get there.
- arca_vorago 10y ago
  I thing the fallacy here is that a single person can't create a large enough dataset, while I content they can. Combined with tools like pocketsphinx I think its very doable to have a privacy concious AI system.
- darkpicnic 10y ago
  This is simply a matter of AI not being advanced enough at the moment. Take how a human learns: it acquires experiences and information over a long childhood, and from this "data set" is able to do millisecond predictions of future events based off completely new data. What needs to happen is a company needs to come along and create AI that is trained off of generalized information... some kind of socially accepted public data set... then the trained core is sold as a seed to individuals who then feed it their personal data. It'll be the equivalent of buying an AI "teenager" and slowly training them to be an "adult".
- internaut 10y ago
  That's the thing. As sensors become richer and the data becomes more valuable to the ML, consumers are becoming more aware of their privacy. That means to get to a 'good seam' in the future instead of trawling through trash, you're going to have to convince millions of people their interests won't be affected. That means in time there is an opportunity for a Google-killer with a different business model not based on using the raw data OR using by the use of intelligent agents. Google goes down because its stakeholders are contingent on getting to the raw data.
JonathonW 10y ago
Google doesn't "spill their beans" to third parties-- what Google actually sells is the opportunity for third parties to be included in the advertisements Google is targeting to their users. Google has a strong incentive to not allow their aggregated user data to leave Google-- the behavioral data Google collects is the reason why Google is valuable; if they start shipping that data off to third parties, suddenly the third parties don't need Google anymore. (Same with Facebook-- they're not "selling" your data; they're selling the opportunity to target you based on your data, but the data itself is too valuable to Facebook to sell.)
- gue5t 10y ago
  Google (or Facebook) is the third party. The first two are you and your computer.
- ProAm 10y ago
  > Google doesn't "spill their beans" to third parties Except the government who gets unfettered access. Not into conspiracy theories but Im definitely not a fan of this (and it goes for all social media and technology companies)
  - Oletros 10y ago
    > Except the government who gets unfettered access. Does you have a source for that?
    - ProAm 10y ago
      >Does you have a source for that? Yep. They all do it. [1] http://news.trust.org/item/20161004170601-99f8c http://news.trust.org/item/20161004170601-99f8c
      - Oletros 10y ago
        You have put a link about Yahoo! and that link says nothing about Google or others doing the same

I think we just envisioned a highly anthropomorphized AI: essentially, a very smart and entirely obedient person to serve as the perfect aide. The Star Trek dream emerged before computing technology was very far advanced, and well before the idea of constantly mobile wireless communication. We thought our AIs would be small and physical, easy for a single person to entirely own and unable to remember more details than a human; instead we got unfamiliar algorithms run on machines far away.

dragonwriter 10y ago
> The Star Trek dream emerged before computing technology was very far advanced, and well before the idea of constantly mobile wireless communication. We thought our AIs would be small and physical, easy for a single person to entirely own and unable to remember more details than a human; instead we got unfamiliar algorithms run on machines far away. Which is kind of funny, because even if it might be accessed by personal mobile devices, the Star Trek "library computer" AI was never "small and physical, easy for a single person to entirely own and unable to remember more details than a human", it was an aspect of a large server (or networked cluster, the actual architecture is somewhat vague) that was part of a capital ship or base, had access in the server/cluster to a library of very nearly all generally available knowledge and extensive personal information about both its users and about people with little direct connection (and could reach out across a galactic network to access additional remote information sources to handle requests). "Unfamiliar algorithms run on machines far away" is much more like the source of the "Star Trek dream" than "small and physical, easy for a single person to entirely own and unable to remember more details than a human".
- bbctol 10y ago
  True! I suppose I was thinking of earlier space-age fantasy more than TNG. It was rather odd how easily people could look up details on crewmembers aboard the Enterprise, but I suppose the situation's different on an organized ship.
  - dragonwriter 10y ago
    > It was rather odd how easily people could look up details on crewmembers aboard the Enterprise, but I suppose the situation's different on an organized ship. Not just crewmembers: TNG showed some of the broader implications (both useful and creepy) of the convenience-oriented panopticon, e.g., when Geordi used the Enterprise library computer to construct a simulation from data (including personality profiles) of Dr. Leah Brahms, who later meets him (and encounters the simulation.) (S306 "Booby Trap", S416 "Galaxy's Child")
    - VLM 10y ago
      That was a substantial plot hole where she had no idea someone was simulating her. I suspect a huge part of the panopticon culture would be / is being informed that you're being peeped at. 99% of the time someone asking "Computer locate commander Riker" involved commander Riker knowing all about who's looking for him and why and having a substantial conversation with the requester. I don't recall any plot along the lines of Deanna getting jealous and spamming the computer all night asking where Riker is and he better not be in that cute ensign's bedroom... Because it seems logical the computer would inform him each time and he would eventually tire of the interruption and nature would take its course, WRT his relationship with Deanna. A better analogy would be technically I could walk up to the company president's office and stalk him, but culturally that is so not going to fly and I would have a lot of explaining to do. Merely using the computer instead of walking there in flesh isn't a major cultural shift. However, every single "holodeck creeper" plot line involved the simulated attractive real world woman not knowing she's being simulated until the plot reached maximum spaghetti spilling cringe, which is one of the few Trek panopticon situations where people being spied on did NOT know they were being spied on, which seems very un-trek, although it made for some entertaining stories. An alternative interpretation is I believe over the course of the series every attractive woman on the ship was simulated on the holodeck at least once by at least one lonely guy, and its possible that culturally they just got used to it, although I find that unlikely. People do get conditioned to become used to the weirdest things, so its not out of the realm of possibility. Possibly a culture of what happens in vegas stays in vegas develops and its just the sexism of the TV show that they never showed the women turning the tables on their fellow male crewmen on the holodeck.
- Spoom 10y ago
  I think Star Trek was very prescient here as well: Security is terrible in the Star Trek universe. Nearly every antagonist is able to break into the computer within about a minute.
gvurrdon 10y ago
Isn't it the case that Apple's attempts at similar AI assistance store much of the data on the user's iPhone?

mattmanser 10y ago

It really is the big problem at the moment with the cutting edge of AI. ML relies on large data-sets and if anyone tried to release a personal device it simply wouldn't even work, let alone compete with the mass surveillance google/ms/amazon are bringing to bear. Unless the state-of-the-art in AI suddenly morphs, we seem to be stuck between giving up our privacy or having vaguely intelligent AI. I personally fall heavily on the privacy side of stuff, but I can see the intellectual and commercial appeal of pretending it doesn't matter in order to get there.

arca_vorago 10y ago
I thing the fallacy here is that a single person can't create a large enough dataset, while I content they can. Combined with tools like pocketsphinx I think its very doable to have a privacy concious AI system.
darkpicnic 10y ago
This is simply a matter of AI not being advanced enough at the moment. Take how a human learns: it acquires experiences and information over a long childhood, and from this "data set" is able to do millisecond predictions of future events based off completely new data. What needs to happen is a company needs to come along and create AI that is trained off of generalized information... some kind of socially accepted public data set... then the trained core is sold as a seed to individuals who then feed it their personal data. It'll be the equivalent of buying an AI "teenager" and slowly training them to be an "adult".
internaut 10y ago
That's the thing. As sensors become richer and the data becomes more valuable to the ML, consumers are becoming more aware of their privacy. That means to get to a 'good seam' in the future instead of trawling through trash, you're going to have to convince millions of people their interests won't be affected. That means in time there is an opportunity for a Google-killer with a different business model not based on using the raw data OR using by the use of intelligent agents. Google goes down because its stakeholders are contingent on getting to the raw data.

JonathonW 10y ago

Google doesn't "spill their beans" to third parties-- what Google actually sells is the opportunity for third parties to be included in the advertisements Google is targeting to their users. Google has a strong incentive to not allow their aggregated user data to leave Google-- the behavioral data Google collects is the reason why Google is valuable; if they start shipping that data off to third parties, suddenly the third parties don't need Google anymore. (Same with Facebook-- they're not "selling" your data; they're selling the opportunity to target you based on your data, but the data itself is too valuable to Facebook to sell.)

gue5t 10y ago
Google (or Facebook) is the third party. The first two are you and your computer.
ProAm 10y ago
> Google doesn't "spill their beans" to third parties Except the government who gets unfettered access. Not into conspiracy theories but Im definitely not a fan of this (and it goes for all social media and technology companies)
- Oletros 10y ago
  > Except the government who gets unfettered access. Does you have a source for that?
  - ProAm 10y ago
    >Does you have a source for that? Yep. They all do it. [1] http://news.trust.org/item/20161004170601-99f8c http://news.trust.org/item/20161004170601-99f8c
    - Oletros 10y ago
      You have put a link about Yahoo! and that link says nothing about Google or others doing the same