• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Friend processes a full day of audio context in real time, Avi Schiffmann says

    Schiffmann describes a screenless wearable with lights, haptics and hands-free conversation, plus a battery that can last all day.

    AV
    1 Source, ,

    TLDR

    Avi Schiffmann says Friend has shipped an always-on AI wearable that turns incoming audio into memories it can retrieve later through a timeline-based knowledge graph. He credits its real-time context processing with allowing it to speak up “only at the right time.” In his September 11 account, he said LTE would eventually make the device fully standalone. He also claimed that, as far as he could tell, Friend was the only team to have shipped a mass-manufacturable version of this kind of proactive AI device.

    Combined views

    5.2K

    1 Source, first seen 19d ago

    Combined views

    5.2K

    1 Source, first seen 19d ago

    21 likes
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    19d ago
    first seen 19d ago
    21 likes
    6 comments
    9 saves

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    6 comments
    9 saves

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1 Source

    @AviSchiffmannNow I know many people have thoughts on the direction I’ve taken friend, and it certainly isn’t directed towards the tech scene. At least in its current state. However for all the yap about the new OAI device, I feel people overlook some of the technical aspects that we have already shipped, with friend. It’s taken awhile but we have a system capable of taking in a full days context, entirely via audio, and synthesizing it in real-time as the thoughts & feelings of an embodied entity. All incoming audio gets thought over, and turned into memories that are later queried through a timeline based knowledge graph. This is packaged into a small wearable device, with a battery that can last all day. There is no screen, but there is lighting, haptics, and the ability to speak in conversation hands free. Eventually with LTE the friend will be fully standalone. At any moment, your friend may speak up. But only at the right time, which is only possible due to the realtime context processing. The image is a recent thought a newly created friend had. (This data is only available for my personal staging friend in a custom diagnostics view, not exposed to customers). As far as I can tell, we are the only team that has shipped a mass manufacturable version of this kind of always-on proactive AI device. Just sayin

    1 Source

    @AviSchiffmannNow I know many people have thoughts on the direction I’ve taken friend, and it certainly isn’t directed towards the tech scene. At least in its current state. However for all the yap about the new OAI device, I feel people overlook some of the technical aspects that we have already shipped, with friend. It’s taken awhile but we have a system capable of taking in a full days context, entirely via audio, and synthesizing it in real-time as the thoughts & feelings of an embodied entity. All incoming audio gets thought over, and turned into memories that are later queried through a timeline based knowledge graph. This is packaged into a small wearable device, with a battery that can last all day. There is no screen, but there is lighting, haptics, and the ability to speak in conversation hands free. Eventually with LTE the friend will be fully standalone. At any moment, your friend may speak up. But only at the right time, which is only possible due to the realtime context processing. The image is a recent thought a newly created friend had. (This data is only available for my personal staging friend in a custom diagnostics view, not exposed to customers). As far as I can tell, we are the only team that has shipped a mass manufacturable version of this kind of always-on proactive AI device. Just sayin