Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Sunday July 16 2017, @09:56PM   Printer-friendly
from the may-I-mambu-dogface-to-the-banana-patch? dept.

Bob: "I can can I I everything else."

Alice: "Balls have zero to me to me to me to me to me to me to me to me to."

To you and I, that passage looks like nonsense. But what if I told you this nonsense was the discussion of what might be the most sophisticated negotiation software on the planet? Negotiation software that had learned, and evolved, to get the best deal possible with more speed and efficiency–and perhaps, hidden nuance–than you or I ever could? Because it is.

This conversation occurred between two AI agents developed inside Facebook. At first, they were speaking to each other in plain old English. But then researchers realized they'd made a mistake in programming.

"There was no reward to sticking to English language," says Dhruv Batra, visiting research scientist from Georgia Tech at Facebook AI Research (FAIR). As these two agents competed to get the best deal–a very effective bit of AI vs. AI dogfighting researchers have dubbed a "generative adversarial network"–neither was offered any sort of incentive for speaking as a normal person would. So they began to diverge, eventually rearranging legible words into seemingly nonsensical sentences.

"Agents will drift off understandable language and invent codewords for themselves," says Batra, speaking to a now-predictable phenomenon that Facebook as observed again, and again, and again. "Like if I say 'the' five times, you interpret that to mean I want five copies of this item. This isn't so different from the way communities of humans create shorthands."

Indeed. Humans have developed unique dialects for everything from trading pork bellies on the floor of the Mercantile Exchange to hunting down terrorists as Seal Team Six–simply because humans sometimes perform better by not abiding to normal language conventions. So should we let our software do the same thing? Should we allow AI to evolve its dialects for specific tasks that involve speaking to other AIs? To essentially gossip out of our earshot? Maybe; it offers us the possibility of a more interoperable world, a more perfect place where iPhones talk to refrigerators that talk to your car without a second thought.

The tradeoff is that we, as humanity, would have no clue what those machines were actually saying to one another.

https://www.fastcodesign.com/90132632/ai-is-inventing-its-own-perfect-languages-should-we-let-it

[Reminds me of]: Voynich Manuscript

What are your thoughts on this topic?


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 5, Insightful) by AthanasiusKircher on Monday July 17 2017, @12:00AM (2 children)

    by AthanasiusKircher (5291) on Monday July 17 2017, @12:00AM (#540084) Journal

    From the summary:

    Maybe; it offers us the possibility of a more interoperable world, a more perfect place where iPhones talk to refrigerators that talk to your car without a second thought.

    Eh? That's a "more perfect place"? The Internet of Things has so far been mediocre in most applications at best, generally involving connecting devices to the internet for no apparent reason other than "we can" (and generally because it somehow makes companies more money). At worst, it has introduced so many security nightmare scenarios that I don't even know where to start.

    In what insane logic is a world where such unnecessary IoT communication compounded by indecipherable botlike exchanges between the already problematic IoT devices a "more perfect place"??

    The tradeoff is that we, as humanity, would have no clue what those machines were actually saying to one another.

    No, the tradeoff is that even if these devices function and communicate "on task" as it were, you're still likely to end up with your refrigerator accidentally shutting down your car and disabling it because you forgot to buy milk from your online grocery list. Or some other completely idiotic confluence of IoT weirdness. Add on the security issues endemic to IoT, and the fact that you can't understand that gibberish your devices are muttering is the least of your worries.

    (To be clear: I'm NOT saying all IoT stuff is bad. Just that right now with current implementations a lot of stuff is problematic and often a "solution in search of a problem.")

    Starting Score:    1  point
    Moderation   +3  
       Insightful=2, Interesting=1, Total=3
    Extra 'Insightful' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   5  
  • (Score: 1, Insightful) by Anonymous Coward on Monday July 17 2017, @01:22AM

    by Anonymous Coward on Monday July 17 2017, @01:22AM (#540109)

    > the possibility of a more interoperable world, a more perfect place where iPhones talk to refrigerators that talk to your car without a second thought.

    Came to say the same thing, this must be the exact opposite of "a more perfect place".

    IoT seems to be doing well as part of factory automation, I(industrial)IoT, where it makes sense to connect many small processors/sensors/machines/etc together to control some process.

    I don't live in a factory, and I don't want to.

  • (Score: 0) by Anonymous Coward on Monday July 17 2017, @12:27PM

    by Anonymous Coward on Monday July 17 2017, @12:27PM (#540271)

    In what insane logic is a world where such unnecessary IoT communication compounded by indecipherable botlike exchanges between the already problematic IoT devices a "more perfect place"??

    In marketing logic.